Fin Market Chapter 5

Risk and Return:
Past and
Prologue
5
Chapter
Learning Objectives:
LO5-1 Compute various measures of return on multi-year investments.
LO5-2 Use data on the past performance of stocks and bonds or scenario analysis
to characterize the risk and return features of these investments.
LO5-3 Determine the expected return and risk of portfolios that are constructed
by combining risky assets with risk-free investments in Treasury
bills.
LO5-4 Use the Sharpe ratio to evaluate the investment performance of a portfolio and
provide a guide for capital allocation.
W
hat constitutes a satisfactory judged on its
investment portfolio? Until the contributions to
early 1970s, a reasonable both the
answer expected return
would have been a federally insured bank and the risk of
savings account (a risk-free asset) plus a the entire
risky portfolio of U.S. stocks. Nowadays,
investors have access to a vast array of
assets and can easily construct portfolios
that include foreign stocks and bonds, real
estate, precious metals, and collectibles. Even
more complex strategies may include futures,
options, and other derivatives to insure
portfolios against specified risks.
Clearly every individual security must be
portfolio. We begin with an examination of various conventions for We then consider the trade-offs that arise
measuring and report- ing rates of return. Next, we turn to the historical when investors practice the simplest form of
performance of several broadly diversified investment portfolios. In doing risk control, capital allocation: choosing the
so, we use a risk-free portfolio of Treasury bills as a benchmark to fraction of the portfolio invested in virtually
evaluate the historical performance of diversified stock and bond risk-free securities versus risky securities.
portfolios. We show how to calculate the performance
one may expect from various allocations
110
between a risk-free asset and a risky can evaluate a passive strategy that will Related websites
for this chapter
portfolio and contemplate the mix that would serve as a benchmark for the active
are available at
best suit different investors. With this strategies considered in the next chapter. www.mhhe.com/bkm.
background, we
5.1 RATES OF RETURN

A key measure of investors’ success is the rate at which their funds have grown during the
investment period. The total holding-period return (HPR) of a share of stock depends on
holding-period return
the increase (or decrease) in the price of the share over the investment period as well as on (HPR)
any dividend income the share has provided. The rate of return is defined as dollars earned
Rate of return over a given
over the investment period (price appreciation as well as dividends) per dollar invested:
investment period.
Ending price 2 Beginning price 1 Cash dividend

HPR 5 Beginning price (5.1)
This definition of the HPR assumes that the dividend is paid at the end of the holding
period. When dividends are received earlier, the definition ignores reinvestment income
between the receipt of the dividend and the end of the holding period. The percentage return
from dividends, cash dividends/beginning price, is called the dividend yield, and so the divi-
dend yield plus the capital gains yield equals the HPR.
This definition of holding return is easy to modify for other types of investments.
For example, the HPR on a bond would be calculated using the same formula, except
that the bond’s interest or coupon payments would take the place of the stock’s divi-
dend payments.
Consider investing some of your money, now all invested in a bank account, in a stock market EXAMPLE 5.1
index fund. The price of a share in the fund is currently $100, and your time horizon is one year.
You expect the cash dividend during the year to be $4, so your expected dividend yield is 4%. Holding-Period Return
Your HPR will depend on the price one year from now. Suppose your best guess is that it will
be
$110 per share. Then your capital gain will be $10, so your capital gains yield is $10/$100 5 .10,
or 10%. The total holding-period rate of return is the sum of the dividend yield plus the capital
gains yield, 4% 1 10% 5 14%.
$110 2 $100 1 $4
HPR 5 5 .14, or 14%
$100
Measuring Investment Returns over Multiple Periods

The holding-period return is a simple and unambiguous measure of investment return over a
single period. But often you will be interested in average returns over longer periods of time.
For example, you might want to measure how well a mutual fund has performed over the
preceding five-year period. In this case, return measurement is more ambiguous.
Consider a fund that starts with $1 million under management. It receives additional
funds from new and existing shareholders and also redeems shares of existing shareholders
so that net cash inflow can be positive or negative. The fund’s quarterly results are as given
in Table 5.1, with negative numbers in parentheses.
The numbers indicate that when the firm does well (i.e., achieves a high HPR), it
attracts new funds; otherwise it may suffer a net outflow. For example, the 10% return
in the first quarter by itself increased assets under management by .10 3 $1 million 5
$100,000; it also elicited new investments of $100,000, thus bringing assets under
management to $1.2 million
111
Chapter 5 Risk and Return: Past and Prologue 112
TABLE 5.1 Quarterly cash flows and rates of return of a mutual fund
1st 2nd 3rd 4th

Quarter Quarter Quarter Quarter
Assets under management at

start of quarter ($ million) 1.0 1.2 2.0 0.8
Holding-period return (%) 10.0 25.0 (20.0) 20.0
Total assets before net inflows 1.1 1.5 1.6 0.96
Net inflow ($ million)* 0.1 0.5 (0.8) 0.6
Assets under management at
end of quarter ($ million) 1.2 2.0 0.8 1.56
*New investment less redemptions and distributions, all assumed to occur at the end of each quarter.
by the end of the quarter. An even better HPR in the second quarter elicited a larger net
inflow, and the second quarter ended with $2 million under management. However, HPR in
the third quarter was negative, and net inflows were negative.
How would we characterize fund performance over the year, given that the fund
experienced both cash inflows and outflows? There are several candidate measures of
performance, each with its own advantages and shortcomings. These are the arithmetic
average, the geometric average, and the dollar-weighted return. These measures may vary
considerably, so it is important to understand their differences.
arithmetic average Arithmetic average The arithmetic average of the quarterly returns is just the
The sum of returns in each sum of the quarterly returns divided by the number of quarters; in the above example:
period divided by the number (10 1 25 2 20 1 20)/4 5 8.75%. Since this statistic ignores compounding, it does not
of periods. represent an equivalent, single quarterly rate for the year. However, without information
beyond the historical sample, the arithmetic average is the best forecast of performance
for the next quarter.
geometric average Geometric average The geometric average of the quarterly returns is equal to the
The single per-period return single per-period return that would give the same cumulative performance as the sequence
that gives the same of actual returns. We calculate the geometric average by compounding the actual period-by-
cumulative performance as period returns and then finding the per-period rate that will compound to the same final
the sequence of actual value. In our example, the geometric average quarterly return, rG, is defined by:
returns.
(1 1 .10) 3 (1 1 .25) 3 (1 2 .20) 3 (1 1 .20) 5 (1 1 rG)4
The left-hand side of this equation is the compounded year-end value of a $1 investment
earning the four quarterly returns. The right-hand side is the compounded value of a $1
investment earning rG each quarter. We solve for rG:
rG 5 [(1 1 .10) 3 (1 1 .25) 3 (1 2 .20) 3 (1 1 .20)]1/4 2 1 5 .0719, or 7.19% (5.2)
The geometric return is also called a time-weighted average return because it ignores the
quarter-to-quarter variation in funds under management. In fact, an investor will obtain a
larger cumulative return when high returns are earned in periods when larger sums have
been invested and low returns are earned when less money is at risk. In Table 5.1, the higher
returns (25% and 20%) were achieved in quarters 2 and 4, when the fund managed
$1,200,000 and
$800,000, respectively. The lower returns (220% and 10%) occurred when the fund managed
$2,000,000 and $1,000,000, respectively. In this case, better returns were earned when less
money was under management—an unfavorable combination.
Published data on past returns earned by mutual funds actually are required to be time-
weighted returns. The rationale for this practice is that since the fund manager does not have
full control over the amount of assets under management, we should not weight returns in
one period more heavily than those in other periods when assessing “typical” past
performance.
Dollar-weighted RETURN To account for varying amounts under management,

we treat the fund cash flows as we would a capital budgeting problem in corporate finance
and compute the portfolio manager’s internal rate of return (IRR). The initial value of $1
million and the net cash inflows are treated as the cash flows associated with an
investment “project.” The year-end “liquidation value” of the portfolio is the final cash
flow of the project. In our example, the investor’s net cash flows are as follows:
Quarter
0 1 2 3 4
Net cash flow ($ million) 21.0 2.1 2.5 .8 2.6 1 1.56 5 .96
The entry for time 0 reflects the starting contribution of $1 million; the negative entries for
times 1 and 2 are additional net inflows in those quarters, while the positive value for quarter
3 signifies a withdrawal of funds. Finally, the entry for time 4 represents the sum of the final
(negative) cash inflow plus the value of the portfolio at the end of the fourth quarter. The
latter is the value for which the portfolio could have been liquidated at year-end.
The dollar-weighted average return is the internal rate of return of the project, which is
dollar-weighted
3.38%. The IRR is the interest rate that sets the present value of the cash flows realized on
average return
the portfolio (including the $1.56 million for which the portfolio can be liquidated at the end
The internal rate of return
of the year) equal to the initial cost of establishing the portfolio. It therefore is the interest
on an investment.
rate that satisfies the following equation:
2.1 2.5 .8 .96

0 5 21.0 1 1 1 1
(5.3)
1 1 IRR (1 1 IRR)2 (1 1 IRR)3 (1 1 IRR)4
The dollar-weighted return in this example is less than the time-weighted return of 7.19%
because, as we noted, the portfolio returns were higher when less money was under manage-
ment. The difference between the dollar- and time-weighted average return in this case is
quite large.
CONCEPT
A fund begins with $10 million and reports the following three-month results (with negative
c h e ck 5.1
figures in parentheses):
Month
1 2 3
Net inflows (end of month, $ million) 3 5 0
HPR (%) 2 8 (4)
Compute the arithmetic, time-weighted, and dollar-weighted average returns.
Conventions for Annualizing Rates of Return

We’ve seen that there are several ways to compute average rates of return. There also is
some variation in how the mutual fund in our example might annualize its quarterly returns.
Returns on assets with regular cash flows, such as mortgages (with monthly payments) and
bonds (with semiannual coupons), usually are quoted as annual percentage rates, or APRs,
which annualize per-period rates using a simple interest approach, ignoring compound
interest:
APR 5 Per-period rate 3 Periods per year
However, because it ignores compounding, the APR does not equal the rate at which your
invested funds actually grow. This is called the effective annual rate, or EAR. When there
are n compounding periods in the year, we first recover the rate per period as APR/n and
then compound that rate for the number of periods in a year. (For example, n 5 12 for
monthly payment mortgages and n 5 2 for bonds making payments semiannually.)
n APR n
1 1 EAR 5 (1 1 Rate per period) 5 a1 1 b (5.4)
n
Since you can earn the APR each period, after one year (when n periods have passed),
your cumulative return is (1 1 APR/n)n. Note that one needs to know the holding period
when given an APR in order to convert it to an effective rate.
Rearranging Equation 5.4, we can also find APR given EAR:
APR 5 [(1 1 EAR)1/n 2 1] 3 n
The EAR diverges by greater amounts from the APR as n becomes larger (we compound
cash flows more frequently). In the limit, we can envision continuous compounding when n
becomes extremely large in Equation 5.4. With continuous compounding, the relationship
between the APR and EAR becomes
1 1 EAR 5 e APR
or,
equivalently,
APR 5 ln(1 1 EAR)
More generally, the EAR of any investment can be converted to an equivalent continuously
compounded rate, rcc , using the relationship
rcc 5 ln(1 1 EAR) (5.5)
We will return to continuous compounding later in the chapter.
EXAMPLE 5.2 Suppose you buy a $10,000 face value Treasury bill maturing in one month for $9,900. On the
bill’s maturity date, you collect the face value. Since there are no other interest payments, the
Annualizing holding- period return for this one-month investment is
Treasury-Bill Returns
Cash income 1 Price $100
HPR 5 change 5 5 .0101 5
$9,900
Initial price 1.01%
The APR on this investment is therefore 1.01% 3 12 5 12.12%. The effective annual rate is
higher:
1 1 EAR 5 (1.0101)12 5 1.1282
which implies that EAR 5 .1282 5 12.82%.
A warning: Terminology can be loose. Occasionally, annual percentage yield or APY and
even APR are used interchangeably with effective annual rate, and this can lead to
confusion. To avoid error, you must be alert to context.
The difficulties in interpreting rates of return over time do not end here. Two thorny
issues remain: the uncertainty surrounding the investment in question and the effect of
inflation.
5.2 RISK AND RISK PREMIUMS

Any investment involves some degree of uncertainty about future holding-period returns,
and in many cases that uncertainty is considerable. Sources of investment risk range from
macroeconomic fluctuations, to the changing fortunes of various industries, to asset-specific
unexpected developments. Analysis of these multiple sources of risk is presented in Part
Four, “Security Analysis.”
Scenario Analysis and Probability Distributions

When we attempt to quantify risk, we begin with the question: What HPRs are possible, and
how likely are they? A good way to approach this question is to devise a list of possible
economic outcomes, or scenarios, and specify both the likelihood (probability) of each
scenario and the HPR the asset will realize in that scenario. Therefore, this approach is called
scenario analysis
scenario analysis. The list of possible HPRs with associated probabilities is the probability
distribution of HPRs. Consider an investment in a broad portfolio of stocks, say, an index Process of devising a list of
fund, which we will refer to as the “stock market.” A very simple scenario analysis for the possible economic scenarios
and specifying the likelihood
stock market (assuming only four possible scenarios) is illustrated in Spreadsheet 5.1.
of each one, as well as the
The probability distribution lets us derive measurements for both the reward and the risk
HPR that will be realized in
of the investment. The reward from the investment is its expected return, which you can each case.
think of as the average HPR you would earn if you were to repeat an investment in the asset
many times. The expected return also is called the mean of the distribution of HPRs and
probability distribution
often is referred to as the mean return.
List of possible outcomes
To compute the expected return from the data provided, we label scenarios by s and
with associated
denote the HPR in each scenario as r (s), with probability p(s). The expected return, denoted
probabilities.
E(r), is then the weighted average of returns in all possible scenarios, s 5 1, . . . , S, with
weights equal to the probability of that particular scenario.
expected return
S The mean value of the
E (r) 5 a p(s) r (s) (5.6) distribution of HPR.
s 51
Each entry in column D of Spreadsheet 5.1 corresponds to one of the products in the
summation in Equation 5.6. The value in cell D7, which is the sum of these products, is
therefore the expected return. Therefore, E (r) 5 10%.
Because there is risk to the investment, the actual return may be (a lot) more or less than
10%. If a “boom” materializes, the return will be better, 30%, but in a severe recession the
return will be a disappointing 237%. How can we quantify this uncertainty?
The “surprise” return in any scenario is the difference between the actual return and the
expected return. For example, in a boom (scenario 4) the surprise is r(4) 2 E(r) 5 30% 2
10% 5
20%. In a severe recession (scenario 1), the surprise is r (1) 2 E(r) 5 237% 2 10% 5 247%.
SPREADSHEET 5.1
Scenario analysis for the stock market
A B C D E F
1 Column B x Deviation from Column B x
2 Scenario Probability HPR (%) Column C Mean Return Squared Deviation
3 1. Severe recession .05 —37 —1.85 —47.00 110.45 Please visit us at
4 2. Mild recession .25 —11 —2.75 —21.00 110.25
5 3. Normal growth .40 14 5.60 4.00 6.40 www.mhhe.com/bkm
6 4. Boom .30 30 9.00 20.00 120.00
7 Column sums: Expected return = 10.00 Variance = 347.10
8 Square root of variance = Standard deviation (%) = 18.63
Uncertainty surrounding the investment is a function of both the magnitudes and the
probabilities of the possible surprises. To summarize risk with a single number, we define
variance the variance as the expected value of the squared deviation from the mean (the expected
The expected value of the squared “surprise” across scenarios).
squared deviation from the
S
mean.
Var(r) ; s2 a p(s)[r (s) 2 E (r)]
2
(5.7)
5 s 51
We square the deviations because negative deviations would offset positive deviations other-
wise, with the result that the expected deviation from the mean return would necessarily
be zero. Squared deviations are necessarily positive. Squaring (a nonlinear -transformation)
exaggerates large (positive or negative) deviations and deemphasizes small deviations.
Another result of squaring deviations is that the variance has a dimension of percent
squared. To give the measure of risk the same dimension as expected return (%), we use the
standard deviation standard deviation, defined as the square root of the variance:
The square root of the
variance. SD (r) ; s 5 "Var (r) (5.8)
EXAMPLE 5.3
Applying Equation 5.6 to the data in Spreadsheet 5.1, we find that the expected rate of return on
Expected Return and the stock index fund is
Standard Deviation E ( r) 5 .05 3 (237) 1 .25 3 (211) 1 .40 3 14 1 .30 3 30 5 10%
We use Equation 5.7 to find the variance. First we take the difference between the holding-
period return in each scenario and the mean return, then we square that difference, and finally
we multiply by the probability of each scenario. The sum of the probability-weighted squared
deviations is the variance.
s2 5 .05 (237 2 10)2 1 .25 (211 2 10)2 1 .40 (14 2 10)2 1 .30(30 2 10)2 5 347.10
and so the standard deviation is
s 5 "347.10 5 18.63%
Column F of Spreadsheet 5.1 replicates these calculations. Each entry in that column is the
squared deviation from the mean multiplied by the probability of that scenario. The sum of the
probability- weighted squared deviations that appears in cell F7 is the variance, and the square
root of that value is the standard deviation (in cell F8).
The Normal Distribution

The normal distribution is central to the theory and practice of investments. Its familiar bell-
shaped plot is symmetric, with identical values for all three standard measures of “typical”
results: the mean (the expected value discussed earlier), the median (the value above and
below which we expect 50% of the observations), and the mode (the most likely value).
Figure 5.1 illustrates a normal distribution with a mean of 10% and standard deviation
(SD) of 20%. Notice that the probabilities are highest for outcomes near the mean and are
significantly lower for outcomes far from the mean. But what do we mean by an outcome
“far” from the mean? A return 15% below the mean would hardly be noteworthy if typical
volatility were high, for example, if the standard deviation of returns were 20%, but that
same outcome would be highly unusual if the standard deviation were only 5%. For this
reason, it is often useful to think about deviations from the mean in terms of how many
standard deviations they represent. If the standard deviation is 20%, that 15% negative
surprise would be only three-fourths of a standard deviation, unfortunate perhaps but not
uncommon. But if the standard deviation were only 5%, a 15% deviation would be a “three-
sigma event,” and quite rare.
FIGURE 5.1
The normal distribution
with mean return 10% and
standard deviation 20%
68.26%
95.44%
99.74%
–4σ –3σ –2σ –1σ 0 +1σ +2σ +3σ +4σ

–70 –50 –30 –10 10 30 50 70 90
Rate of return (%)
We can transform any normally distributed return, ri, into a “standard deviation score,” by
first subtracting the mean return (to obtain distance from the mean or return “surprise”) and
then dividing by the standard deviation (which enables us to measure distance from the
mean in units of standard deviations).
ri 2 E ( r i )
sr 5 (5.9A)
i
si
This standardized return, which we have denoted sri, is normally distributed with a mean of
zero and a standard deviation of 1. We therefore say that sri is a “standard normal” variable.
Conversely, we can start with a standard normal return, sri, and recover the original return
by multiplying by the standard deviation and adding back the mean return:
ri 5 E(ri) 1 sri 3 si (5.9B)
In fact, this is how we drew Figure 5.1. Start with a standard normal (mean 5 0 and SD
5 1); next, multiply the distance from the mean by the assumed standard deviation of
20%; finally, recenter the mean away from zero by adding 10%. This gives us a normal
variable with mean 10% and standard deviation 20%.
Figure 5.1 shows that when returns are normally distributed, roughly two-thirds (more
precisely, 68.26%) of the observations fall within one standard deviation of the mean, that is,
the probability that any observation in a sample of returns would be no more than one
standard deviation away from the mean is 68.26%. Deviations from the mean of more than
two SDs are even rarer: 95.44% of the observations are expected to lie within this range.
Finally, only 2.6 out of 1,000 observations are expected to deviate from the mean by three or
more SDs.
Two special properties of the normal distribution lead to critical simplifications of invest-
ment management when returns are normally distributed:
1. The return on a portfolio comprising two or more assets whose returns are normally
distributed also will be normally distributed.
2. The normal distribution is completely described by its mean and standard deviation.
No other statistic is needed to learn about the behavior of normally distributed returns.
These two properties in turn imply this far-reaching conclusion:
3. The standard deviation is the appropriate measure of risk for a portfolio of assets with
normally distributed returns. In this case, no other statistic can improve the risk
assessment conveyed by the standard deviation of a portfolio.
Suppose you worry about large investment losses in worst-case scenarios for your
portfolio. You might ask: “How much would I lose in a fairly extreme outcome, for exam-
ple, if my return were in the fifth percentile of the distribution?” You can expect your
investment experience to be worse than this value only 5% of the time and better than this
value at risk (VaR) value 95% of the time. In investments parlance, this cutoff is called the value at risk
Measure of downside risk. (denoted by VaR, to distinguish it from Var, the common notation for variance). A loss-
The worst loss that will be averse investor might desire to limit portfolio VaR, that is, limit the loss corresponding to
suffered with a given a probability of 5%.
probability, often 5%. For normally distributed returns, VaR can be derived from the mean and standard
deviation of the distribution. We calculate it using Excel’s standard normal function
5NORMSINV(0.05). This function computes the fifth percentile of a normal
distribution with a mean of zero and a variance of 1, which turns out to be 21.64485. In
other words, a value that is 1.64485 standard deviations below the mean would
correspond to a VaR of 5%, that is, to the fifth percentile of the distribution.
VaR 5 E(r) 1 (21.64485)s (5.10)
We can obtain this value directly from Excel’s nonstandard normal function 5NORMINV
(.05, E(r), s).
When faced with a sample of actual returns that may not be normally distributed, we must
estimate the VaR directly. The 5% VaR is the fifth-percentile rate of return. For a sample of
100 returns this is straightforward: If the rates are ordered from high to low, count the fifth
observation from the bottom.
Calculating the 5% VaR for samples where 5% of the observations don’t make an integer
requires interpolation. Suppose we have 72 monthly observations so that 5% of the sample is
3.6 observations. We approximate the VaR by going .6 of the distance from the third to
the fourth rate from the bottom. Suppose these rates are 242% and 237%. The
interpolated value for VaR is then 242 1 .6 (42 2 37) 5 239%.
In practice, analysts sometimes compare the historical sample VaR to the VaR implied by
a normal distribution with the same mean and SD as the sample rates. The difference
between these VaR values indicates the deviation of the observed rates from normality.
CONCEPT
c h e ck 5.2 a. The current value of a stock portfolio is $23 million. A financial analyst summarizes
the uncertainty about next year’s holding-period return using the scenario analysis in
the following spreadsheet. What are the annual holding-period returns of the portfolio
in each scenario? Calculate the expected holding-period return, the standard
Please visit us at deviation of returns, and the 5% VaR. What is the VaR of a portfolio with normally
www.mhhe.com/bkm distributed returns with the same mean and standard deviation as this stock? The
spreadsheet is available at the Online Learning Center (go to www.mhhe.com/bkm,
and link to the Chapter 5 material).
A B C D E
Business End-of-Year Value Annual Dividend
1
Conditions Scenario, s Probability, p ($ million) ($ million)
2 High growth 1 .30 35 4.40
3 Normal growth 2 .45 27 4.00
4 No growth 3 .20 15 4.00
5 Recession 4 .05 8 2.00
b. Suppose that the worst three rates of return in a sample of 36 monthly observations are
17%, 25%, and 2%. Estimate the 5% VaR.
Normality over Time

The fact that portfolios of normally distributed assets also are normally distributed greatly
simplifies analysis of risk because standard deviation, a simple-to-calculate number, is the
appropriate risk measure for normally distributed portfolios.
But even if returns are normal for any particular time period, will they also be normal for
other holding periods? Suppose that monthly rates are normally distributed with a mean of
1%. The expected annual rate of return is then 1.0112 2 1. Can this annual rate, which is a
nonlinear function of the monthly return, also be normally distributed? Unfortunately, the
answer is no. Similarly, why would monthly rates be normally distributed when a monthly
rate is (1 1 daily rate)30 2 1? Indeed, they are not. So, do we really get to enjoy the
simplifications offered by the normal distribution?
Despite these potential complications, when returns over very short time periods (e.g.,
an hour or even a day) are normally distributed, then HPRs up to holding periods as long
as a month will be nearly normal, and we can treat them as if they are normal. Longer-
term, for example, annual, HPRs will indeed deviate more substantially from normality,
but even here, if we expressed those HPRs as continuously compounded rates, they will
remain normally distributed. The practical implication is this: Use continuously com-
pounded rates in all work where normality plays a crucial role, as in estimating VaR from
actual returns.
To see why relatively short-term rates are still nearly normal, consider these calculations:
Suppose that rates are normally distributed over an infinitesimally short period. Beyond that,
compounding, strictly speaking, takes them adrift from normality. But those deviations will
be very small. Suppose that on an annual basis the continuously compounded rate of return
has a mean of .12 (i.e., 12%; we must work with decimals when using continuously
compounded rates). Equivalently, the effective annual rate has an expected value of E(r) 5 e
.12
2 1 5 0.1275. So the difference between the effective annual rate and continuously
compounded rate is meaningful, .75%, or 75 basis points. On a monthly basis, however, the
equivalent continuously compounded expected holding-period return is 1%, implying an
expected monthly effective rate of e .01 2 1 5 .01005. The difference between effective annual
and continuously compounded rates here is trivial, only one-half of a basis point. For shorter
periods the difference will be smaller still. So, when continuously compounded rates are
exactly normal, rates over periods up to a month are so close to those continuously
compounded values that we can treat them as if they are effectively normal.
Another important aspect of (normal) continuously compounded rates over time is this:
Just as the total continuously compounded rate and the risk premium grow in direct propor-
tion to the length of the investment period, so does the variance (not the standard deviation)
of the total continuously compounded return and the risk premium. Hence, for an asset with
annual continuously compounded SD of .20 (20%), the variance is .04, and the quarterly
variance will be .01, implying a quarterly standard deviation of .10, or 10%. (Verify that
the monthly standard deviation is 5.77%.) Because variance grows in direct proportion to
time, the standard deviation grows in proportion to the square root of time.
Deviation from Normality and Value at Risk

The scenario analysis laid out in Spreadsheet 5.1 offers insight about the issue of normality
in practice. While a four-scenario analysis is quite simplistic, even this simple example can
nevertheless shed light on how practical analysis might take shape.1
How can the returns specified in the scenario analysis in Spreadsheet 5.1 be judged
against the normal distribution? (As prescribed above, we first convert the effective rates
specified
1
You may wonder: Is the fact that the probability of the worst-case scenario is .05 in Spreadsheet 5.1 just a lucky
happenstance given our interest in the 5% VaR? The answer is no. Given investor concern about VaR, it is fair (in
fact, necessary) to demand of analysts that their scenario analysis explicitly take a stand on the rate of return
corresponding to the probability of the VaR of interest, here .05.
(from Spreadsheet 5.1) to a 3 1.0

normal distribution with the
FIGURE 5.2
A B C D E F G
Comparing scenario analysis 1
2
4
same mean and standard 5 0.9 Scenario analysis
deviation 6
Likewise normal
7 0.8
8
9 0.7
10
11 0.6
12
13 0.5
14
15 0.4
16
17 0.3
18
19 0.2
20
21
0.1
22
23
0
–0.4620 –0.1165 0.1310 0.2624
24
25 Continuously compounded (cc) rates
(decimal)
26
27 Cumulative
28 Scenario Analysis Squared Corresponding
29 Probability Cumulative Effective Rate cc Rate (decimal) Deviations Normal
30 0.05 0.05 –0.37 20.4620 0.2926 0.0020
31 0.25 0.3 –0.11 20.1165 0.0382 0.1494
32 0.40 0.7 0.14 0.1310 0.0027 0.6092
33 0.30 1 0.30 0.2624 0.0337 0.8354
34 Mean 0.10 0.0789
35 SD 0.1881
36
37 VaR Scenario Analysis –0.37 20.4620
38 VaR corresponding Normal –0.2058 20.2304
39
in each scenario to their equivalent continuously compounded rates using Equation 5.5.)
Obviously, it is naive to believe that this simple analysis includes all possible rates. But
while we cannot explicitly pin down probabilities of rates other than those given in the
table, we can get a good sense of the entire spectrum of potential outcomes by
examining the distribution of the assumed scenario rates, as well as their mean and
standard deviation.
Figure 5.2 shows the known points from the cumulative distribution of the
scenario analysis next to the corresponding points from a “likewise normal
distribution” (a normal distribution with the same mean and standard deviation, SD).
Below the graph, we see a table of the actual distributions. The mean in cell D34 is
computed from the formula 5SUMPRODUCT($B$30:$B$33, D30:D33), where the
probability cells B30:B33 are fixed to allow copying to the right.2 Similarly, the SD in
cell F35 is computed from 5SUMPRODUCT(B30:B33, F30:F33)^0.5. The 5% VaR of
the normal distribution in cell E38 is computed from 5NORMINV(0.05, E34, F35).
VaR values appear in cells D37 and D38. The VaR from the scenario analysis, 237%,
is far worse than the VaR derived from the corresponding normal distribution,
220.58%. This immediately suggests that the scenario analysis entails a higher probability
of extreme losses than would be consistent with a normal distribution. On the other hand,
the normal distribution allows for the possibility of extremely large returns, beyond the
maximum return of 30% envisioned in the scenario analysis. We conclude that the
scenario analysis has a distribution that is skewed to the left compared to the normal. It
has a longer left tail (larger losses) and a
2
The Excel function SUMPRODUCT multiplies each term in the first column specified (in this case, the
probabilities in column B) with the corresponding terms in the second column specified (in this case, the
returns in column D), and then adds up those products. This gives us the expected rate of return across
scenarios.
shorter right tail (smaller gains). It makes up for this negative attribute with a larger
probability of positive, but not extremely large, gains (14% and 30%).
This example shows when and why the VaR is an important statistic. When returns are nor-
mal, knowing just the mean and standard deviation allows us to fully describe the entire
distribution. In that case, we do not need to estimate VaR explicitly—we can calculate it
exactly from the properties of the normal distribution. But when returns are not normal, the
VaR conveys important additional information beyond mean and standard deviation. It gives
us additional insight into the shape of the distribution, for example, skewness or risk of
extreme negative outcomes.3
Because risk is largely driven by the likelihood of extreme negative returns, two
additional statistics are used to indicate whether a portfolio’s probability distribution differs
kurtosis
significantly from normality with respect to potential extreme values. The first is kurtosis,
Measure of the fatness of
which compares the frequency of extreme values to that of the normal distribution. The
the tails of a probability
kurtosis of the normal distribution is zero, so positive values indicate higher frequency of
distribution relative to that
extreme values than this benchmark. A negative value suggests that extreme values are less
of a normal distribution.
frequent than with the normal distribution. Kurtosis sometimes is called “fat tail risk,” as Indicates likelihood of
plots of probability distributions with higher likelihood of extreme events will be higher extreme outcomes.
than the normal distribution at both ends or “tails” of the distribution; in other words, the
distributions exhibit “fat tails.” Similarly, exposure to extreme events is often called tail
risk, because these are outcomes in the far reaches or “tail” of the probability distribution.
skew
The second statistic is the skew, which measures the asymmetry of the distribution. Skew
takes on a value of zero if, like the normal, the distribution is symmetric. Negative skew Measure of the asymmetry of
a probability distribution.
suggests that extreme negative values are more frequent than extreme positive ones.
Nonzero values for kurtosis and skew indicate that special attention should be paid to the
VaR, in addition to the use of standard deviation as measure of portfolio risk.
Using Time Series of Return

Scenario analysis postulates a probability distribution of future returns. But where do the
probabilities and rates of return come from? In large part, they come from observing a
sample history of returns. Suppose we observe a 10-year time series of monthly returns on a
diversified portfolio of stocks. We can interpret each of the 120 observations as one potential
“scenario” offered to us by history. Adding judgment to this history, we can develop a
scenario analysis of future returns. As a first step, we estimate the expected return,
standard deviation, and VaR for the sample history. We assume that each of the 120
returns represents one independent draw from the historical probability distribution.
Hence, each return is assigned an equal probability of 1/120 5 .0083. When you use a fixed
probability in Equation 5.6, you obtain the simple
average of the observations, often used to estimate the mean return.
As mentioned earlier, the same principle applies to the VaR. We sort the returns from
high to low. The bottom six observations comprise the lower 5% of the distribution. The
sixth observation from the bottom is just at the fifth percentile, and so would be the 5% VaR
for the historical sample.
Estimating variance from Equation 5.7 requires a minor correction. Remember that vari-
ance is the expected value of squared deviations from the mean return. But the true mean is
not observable; we estimate it using the sample average. If we compute variance as the
average of squared deviations from the sample average, we will slightly underestimate it
because this procedure ignores the fact that the average necessarily includes some estimation
error. The necessary correction turns out to be simple: With a sample of n observations, we
divide the sum of the squared deviations from the sample average by n 2 1 instead of n.
Thus, the estimates of variance and standard deviation from a time series of returns, rt, are
1 1
Var(rt ) 5 S(rt 2 rt )2 SD(rt ) 5 "Var(rt ) rt 5 Srt (5.11)
n21 n
3
The financial crisis of 2008–2009 demonstrated that bank portfolio returns are far from normally distributed, with
exposure to unlikely but catastrophic returns in extreme market meltdowns. The international Basel accord on bank
regulation requires banks to monitor portfolio VaR to better control risk.
EXAMPLE 5.4
To illustrate how to calculate average returns and standard deviations from historical data, let’s
Historical Means and compute these statistics for the returns on the S&P 500 portfolio using five years of data from the
Standard Deviations following table. The average return over this period is 16.7%, computed by dividing the sum of
column (1), below, by the number of observations. In column (2), we take the deviation of each
year’s return from the 16.7% average return. In column (3), we calculate the squared deviation. The
variance is, from Equation 5.11, the sum of the five squared deviations divided by (5 2 1). The
standard deviation is the square root of the variance. If you input the column of rates into a
spreadsheet, the AVERAGE and STDEV functions will give you the statistics directly.
(2) (3)
(1) Deviation from Squared
Year Rate of Return Average Deviation
Return
1 16.9% 0.2% 0.0
2 31.3 14.6 213.2
3 23.2 219.9 396.0
4 30.7 14.0 196.0
5 7.7 29.0 81.0
Total 83.4% 886.2
Average rate of return 5 83.4/5 5 16.7

1
Variance 5
5 2 1 3 886.2 5 221.6
Standard deviation 5 "221.6 5 14.9%
Risk Premiums and Risk Aversion

How much, if anything, would you invest in the index stock fund described in Spreadsheet
5.1? First, you must ask how much of an expected reward is offered to compensate for the
risk involved in stocks.
We measure the “reward” as the difference between the expected HPR on the index fund
risk-free rate and the risk-free rate, the rate you can earn on Treasury bills. We call this difference the
The rate of return that can be risk premium. If the risk-free rate in the example is 4% per year, and the expected index
earned with certainty. fund return is 10%, then the risk premium on stocks is 6% per year.
The rate of return on Treasury bills also varies over time. However, we know the rate of
risk premium return on T-bills at the beginning of the holding period, while we can’t know the return we
will earn on risky assets until the end of the holding period. Therefore, to study the risk
An expected return in excess
of that on risk-free securities.
premium on risky assets we compile a series of excess returns, that is, returns in excess of the
T-bill rate in each period. A reasonable forecast of an asset’s risk premium is the average of its
excess return
historical excess returns.
The degree to which investors are willing to commit funds to stocks depends on risk
Rate of return in excess of
aversion. It seems obvious that investors are risk averse in the sense that, without a positive
the risk-free rate.
risk premium, they would not be willing to invest in stocks. In theory then, there must
always be a positive risk premium on all risky assets in order to induce risk-averse investors
risk aversion
to hold the existing supply of these assets.
Reluctance to accept risk. A positive risk premium distinguishes speculation from gambling. Investors taking on
risk to earn a risk premium are speculating. Speculation is undertaken despite the risk
because of a favorable risk-return trade-off. In contrast, gambling is the assumption of risk
for no purpose beyond the enjoyment of the risk itself. Gamblers take on risk even without a
risk premium.4
4
Sometimes a gamble might seem like speculation to the participants. If two investors differ in their forecasts of the
future, they might take opposite positions in a security, and both may have an expectation of earning a positive risk
premium. In such cases, only one party can, in fact, be correct.
To determine an investor’s optimal portfolio strategy, we need to quantify his degree of

risk aversion. To do so, we look at how he is willing to trade off risk against expected
return. An obvious benchmark is the risk-free asset, which has neither volatility nor risk
premium: It pays a certain rate of return, rf . Risk-averse investors will not hold risky
assets without the prospect of earning some premium above the risk-free rate. An
individual’s degree of risk aversion can be inferred by contrasting the risk premium on
the investor’s entire wealth (the complete portfolio, C ), E(rC) 2 rf , against the variance of
the portfolio return, s2C. Notice that the risk premium and the level of risk that can
be attributed to individual assets in the complete wealth portfolio are of no concern to the
investor here. All that counts is the bottom line: complete portfolio risk premium versus
complete portfolio risk.
A natural way to proceed is to measure risk aversion by the risk premium necessary to
compensate an investor for investing his entire wealth in a portfolio, say Q, with a variance,
s2Q. This approach relies on the principle of revealed preference: We infer preferences from
the choices individuals are willing to make. We will measure risk aversion by the risk
premium offered by the complete portfolio per unit of variance. This ratio measures the
compensation that an investor has apparently required (per unit of variance) to be induced to
hold this portfolio. For example, if we were to observe that the entire wealth of an investor
is held in a portfolio with annual risk premium of .10 (10%) and variance of .0256 (SD 5
16%), we would infer this investor’s degree of risk aversion as:
E ( r Q ) 2 rf 0.10
A5 5 5 3.91 (5.12)
sQ2 0.0256
We call the ratio of a portfolio’s risk premium to its variance the price of risk.5
Later in the section, we turn the question around and ask how an investor with a given price of risk
degree of risk aversion, say, A 5 3.91, should allocate wealth between the risky and risk-free The ratio of portfolio risk
assets. premium to variance.
To get an idea of the level of the risk aversion exhibited by investors in U.S. capital
markets, we can look at a representative portfolio held by these investors. Assume that all
short-term borrowing offsets lending; that is, average borrowing/lending is zero. In that case,
the average investor holds a complete portfolio represented by a stock-market index; 6 call it
M. A common proxy for the market index is the S&P 500 Index. Using a long-term series
of historical returns on the S&P 500 to estimate investors’ expectations about mean return
and variance, we can recast Equation 5.12 with these stock market data to obtain an estimate
of average risk aversion:
Average(rM) 2 rf 0.08
A5 Sample sM2 < 0.045 2 (5.13)
The price of risk of the market index portfolio, which reflects the risk aversion of the
average investor, is sometimes called the market price of risk. Conventional wisdom holds
that plausible estimates for the value of A lie in the range of 1.524. (Take a look at average
excess returns and SD of the stock portfolios in Table 5.2, and compute the risk aversion
implied by their histories to investors that invested in them their entire wealth.)
The Sharpe (Reward-to-Volatility) Ratio

Risk aversion implies that investors will accept a lower reward (as measured by their portfolio
risk premium) in exchange for a sufficient reduction in the standard deviation. A statistic
5
Notice that when we use variance rather than the SD, the price of risk of a portfolio does not depend on the
holding period. The reason is that variance is proportional to the holding period. Since portfolio return and risk
premium also are proportional to the holding period, the portfolio pays the same price of risk for any holding
period.
6
In practice, a broad market index such as the S&P 500 often is taken as representative of the entire market.
TABLE 5.2 Annual rate-of-return statistics for diversified portfolios for 1926–2010 and three subperiods
(%)
World Portfolio U.S. Market
Equity Return Bond Return Long-Term

in U.S. Dollars in U.S. Dollars Small Stocks Large Stocks T-Bonds
Total Return—Geometric Average

1926–2010 9.21 5.42 11.80 9.62 5.12
1926–1955 8.31 2.54 11.32 9.66 3.46
1956–1985 10.28 5.94 13.81 9.52 4.64
1986–2010 9.00 8.34 9.99 9.71 7.74
Total Real Return—Geometric Average
1926–2010 6.03 2.35 8.54 6.43 2.06
1926–1955 6.86 1.16 9.82 8.18 2.07
1956–1985 5.23 1.09 8.60 4.51 20.15
1986–2010 5.99 5.36 6.96 6.68 4.77
Excess Return Statistics
Arithmetic average
1926–2010 7.22 2.09 13.91 8.00 1.76
1926–1955 9.30 1.75 20.02 11.67 2.43
1956–1985 5.55 0.38 12.18 5.01 20.87
1986–2010 6.74 4.54 8.66 7.19 4.11
Standard deviation
1926–2010 18.98 8.50 37.56 20.70 7.93
1926–1955 21.50 8.10 49.25 25.40 4.12
1956–1985 16.33 8.42 32.31 17.58 8.29
1986–2010 19.27 8.81 25.82 17.83 10.07
Minimum (lowest excess return)
1926–2010 241.97 218.50 255.34 246.65 213.43
1926–1955 241.03 213.86 255.34 246.65 26.40
1956–1985 232.49 218.50 245.26 234.41 213.09
1986–2010 241.97 211.15 241.47 238.44 213.43
Maximum (highest excess return)
1926–2010 70.51 28.96 152.88 54.26 26.07
1926–1955 70.51 28.96 152.88 54.26 10.94
1956–1985 35.25 26.40 99.94 42.25 24.96
1986–2010 36.64 24.40 73.73 32.11 26.07
Deviation from the Normal Distribution*
Kurtosis
1926–2010 1.49 1.01 0.65 1.05 0.24
1926–1955 1.88 3.05 0.03 0.97 20.24
1956–1985 0.25 1.52 20.08 0.04 0.99
1986–2010 1.85 20.31 0.53 1.93 20.45
Skew
1926–2010 20.83 0.44 20.40 20.86 0.16
1926–1955 20.67 0.64 20.49 21.01 20.20
1956–1985 20.61 0.44 20.31 20.52 0.79
1986–2010 21.36 0.26 20.45 21.30 20.26
Performance Statistics
Sharpe ratio
1926–2010 0.38 0.25 0.37 0.39 0.22
1926–1955 0.43 0.22 0.41 0.46 0.59
(continued)
TABLE 5.2 (concluded)
World Portfolio U.S. Market
Equity Return Bond Return Long-Term

in U.S. Dollars in U.S. Dollars Small Stocks Large Stocks T-Bonds
1956–1985 0.34 0.05 0.38 0.28 20.11

1986–2010 0.35 0.51 0.34 0.40 0.41
VaR*
1926–2010 227.41 210.81 265.13 236.86 211.69
1926–1955 240.04 214.55 278.60 253.43 25.48
1956–1985 229.08 213.53 249.53 230.51 212.46
1986–2010 246.35 210.25 249.16 242.28 213.85
Difference of actual VaR from VaR of a Normal distribution with same mean and SD
1926–2010 22.62 0.34 218.22 29.40 20.99
1926–1955 213.58 23.32 216.51 220.34 21.22
1956–1985 28.19 21.15 210.38 26.89 1.16
1986–2010 218.66 21.03 215.33 218.26 21.83
*Applied to continuously compounded (cc) excess returns (5 cc total return 2 cc T-bill rates).
Source: Inflation data: BLS; T-bills and U.S. small stocks: Fama and French, http://mba.tuck.dart mouth.edu/pages/faculty/ken.
french/data_library.html; Large U.S. stocks: S&P500; Long-term U.S. government bonds: 1926–2003 return on 20-Year U.S. Please visit us at
Treasury bonds, and 2004–2008 Lehman Brothers long-term Treasury index; World portfolio of large stocks: Datastream; World www.mhhe.com/bkm
portfolio of Treasury bonds: 1926–2003 Dimson, Elroy, and Marsh, and 2004–2008 Datastream.
commonly used to rank portfolios in terms of this risk-return trade-off is the Sharpe (or
reward-to-volatility) ratio, defined as Sharpe (or reward-to-
volatility) ratio
Ratio of portfolio risk
Portfolio risk premium E(rP) 2 rf premium to standard
S5 5 deviation.
Standard deviation of portfolio excess (5.14)
sP
return
A risk-free asset would have a risk premium of zero and a standard deviation of zero.
Therefore, the reward-to-volatility ratio of a risky portfolio quantifies the incremental reward
(the increase in risk premium) for each increase of 1% in the portfolio standard deviation.
For example, the Sharpe ratio of a portfolio with an annual risk premium of 8% and standard
deviation of 20% is 8/20 5 0.4. A higher Sharpe ratio indicates a better reward per unit of
volatility, in other words, a more efficient portfolio. Portfolio analysis in terms of mean and
standard deviation (or variance) of excess returns is called mean- variance analysis.
A warning: We will see in the next chapter that while standard deviation and VaR of mean-variance analysis
returns are useful risk measures for diversified portfolios, these are not useful ways to think Ranking portfolios by their
about the risk of individual securities. Therefore, the Sharpe ratio is a valid statistic only for Sharpe ratios.
ranking portfolios; it is not valid for individual assets. For now, therefore, let’s examine the
historical reward-to-volatility ratios of broadly diversified portfolios that reflect the
performance of some important asset classes.
CONCEPT
a. A respected analyst forecasts that the return of the S&P 500 Index portfolio over the
c h e ck 5.3
coming year will be 10%. The one-year T-bill rate is 5%. Examination of recent returns
of the S&P 500 Index suggests that the standard deviation of returns will be 18%. What
does this information suggest about the degree of risk aversion of the average investor,
assuming that the average portfolio resembles the S&P 500?
b. What is the Sharpe ratio of the portfolio in (a)?
5.3 THE HISTORICAL RECORD

World and U.S. Risky Stock and Bond Portfolios
We begin our examination of risk with an analysis of a long sample of return history (85
years) for five risky asset classes. These include three well-diversified stock portfolios—
world large stocks, U.S. large stocks, and U.S. small stocks—as well as two long-term bond
portfolios— world and U.S. Treasury bonds. The 85 annual observations for each of the five
time series of returns span the period 1926–2010.
Until 1969, the “World Portfolio” of stocks was constructed from a diversified sample of
large capitalization stocks of 16 developed countries weighted in proportion to the relative
size of gross domestic product. Since 1970 this portfolio has been diversified across 24
developed countries (almost 6,000 stocks) with weights determined by the relative
capitalization of each market. “Large Stocks” is the Standard & Poor’s market value–
weighted portfolio of 500 U.S. common stocks selected from the largest market
capitalization stocks. “Small U.S. Stocks” are the smallest 20% of all stocks trading on the
NYSE, NASDAQ , and Amex (currently almost 1,000 stocks).
The World Portfolio of bonds was constructed from the same set of countries as the
World Portfolio of stocks, using long-term bonds from each country. Until 1996, “Long-
Term T-Bonds” were represented by U.S. government bonds with at least a 20-year maturity
and approximately current-level coupon rate.7 Since 1996, this bond series has been
measured by the Barclay’s (formerly the Lehman Brothers) Long-Term Treasury Bond
Index.
Look first at Figure 5.3, which shows histograms of total (continuously compounded)
returns of the five risky portfolios and of Treasury bills. Notice the hierarchy of risk: Small
stocks are the most risky, followed by large stocks and then long-term bonds. At the same
time, the higher average return offered by riskier assets is evident, consistent with investor
risk aversion. T-bill returns are by far the least volatile. In fact, despite the variability in their
returns, bills are actually riskless, since you know the return you will earn at the beginning
of the holding period. The small dispersion in these returns reflects the variation in interest
rates over time.
Figure 5.4 provides another view of the hierarchy of risk. Here we plot the year-by-year
returns on U.S. large stocks, long-term Treasury bonds, and T-bills. Risk is reflected by
wider swings of returns from year to year.
Table 5.2 presents statistics of the return history of the five portfolios over the full 85-year
period, 1926–2010, as well as for three subperiods. 8 The first 30-year subperiod, 1926–1955,
includes the Great Depression (1929–1939), World War II, the postwar boom, and a
subsequent recession. The second subperiod (1956–1985) includes four recessions (1957–
1958, 1960–1961, 1973–1975, and 1980–1982) and a period of “stagflation” (poor growth
combined with high inflation (1974–1980). Finally, the most recent 25-year subperiod (1986–
2010) included two recessions (1990–1991, 2001–2003) bracketing the so-called high-tech
bubble of the 1990s, and a severe recession that started in December 2007 and is estimated to
have ended in the second half of 2009. Let us compare capital asset returns in these three
subperiods.
We start with the geometric averages of total returns in the top panel of the table. This is
the equivalent, constant annual rate of return that an investor would have earned over the
period. To appreciate these rates, you must consider the power of compounding. Think about
an investor who might have chosen to invest in either large U.S. stocks or U.S. long-term T-
bonds at the end of 1985. The geometric averages for 1986–2010 tell us that over the most
recent 25-year period, the stock portfolio would have turned $1 into $1 3 1.097125 5
$10.13, while the same investment in the T-bond portfolio would have brought in
$1 3 1.077425 5 $6.45. We will see later that T-bills would have provided only $2.74.
7
The importance of the coupon rate when comparing returns on bonds is discussed in Part Three.
8
Year-by-year returns are available on the Online Learning Center. Go to www.mhhe.com/bkm, and link to
material for Chapter 5.
FIGURE 5.3
Frequency distribution of annual, continuously compounded rates of return, 1926–2010
Source: Prepared from data used in Table 5.2.
T-Bills
0.70
Average return = 3.56
0.60 Standard deviation = 2.95
0.50
World Equity Portfolio (MSCI World)
0.40 0.40
Standard deviation = 18.06
0.30 0.30
0.20 0.20
0.10 0.10
0.00 0.00
–90 –60 –30 0 30 60 90 –90 –60 –30 0 30 60 90
U.S. Large Stocks (S&P 500) Long-Term U.S. Treasury

0.40 0.40 Bonds
Average return = 9.19 Average return = 4.99
0.30 Standard deviation = 19.96 0.30 Standard deviation = 7.58
0.20 0.20
0.10 0.10
0.00 0.00
–90 –60 –30 0 30 60 90 –90 –60 –30 0 30 60 90
U.S. Small Stocks (Lowest Capitalization Quintile) World Bond Portfolio (Barclay’s World
0.40 0.40
Treasuries)
0.30 Standard deviation = 32.72 0.30 Standard deviation = 7.92
0.20 0.20
0.10 0.10
0.00 0.00
–90 –60 –30 0 30 60 90 –90 –60 –30 0 30 60 90
FIGURE 5.4
60 Rates of return on stocks,
bonds, and bills, 1926–2010
40 Source: Prepared from data
used in Table 5.2.
Rate of return (%)
20
220
Large stocks
240 Long-term bonds
T-bills
260
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
Thus, while the differences in average returns in Table 5.2 may seem modest at first glance,
they imply great differences in long-term results. Naturally, the reason all investors don’t
invest everything in stocks is the higher risk that strategy would entail.
The geometric average is always less than the arithmetic average. For a normal
distribution, the difference is exactly half the variance of the return (with returns measured
as decimals, not percentages). Here are the arithmetic averages (from Figure 5.3) and
geometric averages (from Table 5.2) for the three stock portfolios over the period (1926–
2010), the differences between the two averages, as well as half the variance computed from
the respective standard deviations.
Average Portfolio Return (%)
World Stocks U.S. Small Stocks U.S. Large Stocks
Arithmetic average 10.89 17.57 11.67

Geometric average 9.21 11.80 9.62
Difference 1.68 5.78 2.04
Half historical variance 1.75 6.84 2.09
You can see that the differences between the geometric and arithmetic averages are
conse- quential and generally close to one-half the variance of returns, suggesting that these
distributions may be approximately normal, but there is a greater discrepancy for small
stocks; therefore, VaR will still add important information about risk beyond standard
deviation, at least for this asset class.
We have suggested that the geometric average is the correct measure for historical per-
spective. But investors are concerned about their real (inflation-adjusted) rates of return, not
the paper profits indicated by the nominal (dollar) return. The real geometric averages sug-
gest that the real cost of equity capital for large corporations has been about 6%. Notice
from Table 5.2 that the average real rate on small stocks has been consistently declining,
steadily approaching that of large stocks. One reason is that the average size of small, pub-
licly traded firms has grown tremendously. Although they are still far smaller than the larger
firms, their size apparently has reached the level where there is little remaining small-firm
premium. The higher-than-historical-average returns recently provided by long-term bonds
are due largely to capital gains earned as interest rates plunged in the recessions of the
decade ending in 2010.
In the previous section we discussed the importance of risk and risk premiums. Let us
now turn to the excess-return panel of Table 5.2. Notice first that excess returns do not need
to be adjusted for inflation because they are returns over and above the nominal risk-free
rate. Second, bond portfolios, albeit an important asset class, are not really candidates for an
investor’s sole-investment vehicle, because they are not sufficiently diversified. Third, the
large differences in average returns across historical periods reflect the tremendous volatility
of annual returns. One might wonder whether the differences across subperiods are
statistically significant. Recalling that the standard deviation of the average return is the
annual standard deviation divided by the square root of the number of observations, none of
the differences between these subperiod averages and the 1926–2010 average exceeds one
standard deviation for stocks and 1.8 standard deviations for bonds. Thus, differences in
these subperiod results might well reflect no more than statistical noise.
The minimum and maximum historical returns also reflect the large variability in
annual returns. Notice the large worst-case annual losses (around 50%) and even larger
best-case gains (50%–150%) on the stock portfolios, as well as the more moderate extreme
returns on the bond portfolios. Interestingly, the small and large U.S. stock portfolios
each experienced both their maximum and minimum returns during the Great Depres-
sion; indeed, that period is also associated with the largest standard deviations of stock
portfolio returns.
The potential import of the risk premium can be illustrated with a simple example. Consider two EXAMPLE 5.5
investors with $1 million as of December 31, 2000. One invests in the small-stock portfolio, and
the other in T-bills. Suppose both investors reinvest all income from their portfolios and liquidate The Risk Premium
their investments 10 years later, on December 31, 2010. We can find the annual rates of return and Growth of
for this period from the spreadsheet of returns at the Online Learning Center. (Go to Wealth
www.mhhe.com/bkm. Look for the link to Chapter 5 material.) We compute a “wealth index” for
each investment by compounding wealth at the end of each year by the return earned in the
following year. For example, we calculate the value of the wealth index for small stocks as of
2003 by multiplying the value as of 2002 (1.1404) by 1 plus the rate of return earned in 2003
(measured in decimals), that is, by 1 1 .7475, to obtain 1.9928.
Small Stocks T-Bills

Year Return (%) Wealth Index Return (%) Wealth Index
2000 1 1
2001 29.25 1.2925 3.86 1.0386
2002 211.77 1.1404 1.63 1.0555
2003 74.75 1.9928 1.02 1.0663
2004 14.36 2.2790 1.19 1.0790
2005 3.26 2.3533 2.98 1.1111
2006 17.69 2.7696 4.81 1.1646
2007 28.26 2.5408 4.67 1.2190
2008 239.83 1.5288 1.64 1.2390
2009 36.33 2.0842 0.05 1.2396
2010 29.71 2.7034 0.08 1.2406
The final value of each portfolio as of December 31, 2010, equals its initial value ($1 million)
multiplied by the wealth index at the end of the period:
Date Small Stocks T-Bills
December 31, 2000 $1,000,000 $1,000,000

December 31, 2010 $2,703,420 $1,240,572
The difference in total return is dramatic. Even with its devasting 2008 return, the value of the
small-stock portfolio after 10 years is 118% more than that of the T-bill portfolio.
We can also calculate the geometric average return of each portfolio over this period. For T-
bills, the geometric average over the 10-year period is computed from:
(1 1 rG )10 5 1.2406
1 1 rG 5 1.24061/10 5 1.0218
rG 5 2.18%
Similarly, the geometric average for small stocks is 10.46%. The difference in geometric average
reflects the difference in cumulative wealth provided by the small-stock portfolio over this period.
Are these portfolios normally distributed? The next section of Table 5.2 shows the kurto-
sis and skew of the distributions. As discussed earlier, testing for normality requires us to
use continuously compounded rates. Accordingly, we use Equation 5.5 to compute
continuously compounded rates of return. We calculate ln(1 1 annual rate) for each asset and
compute excess returns by subtracting the continuously compounded rate of return on T-
bills. Because
these measures derive from higher exponents of deviations from the mean (the cubed devia-
tion for skew and the fourth power of the deviation for kurtosis), these measures are highly
sensitive to rare but extreme outliers; therefore, we can rely on these measures only in very
large samples that allow for sufficient observations to be taken as exhibiting a “representa-
tive” number of such events. You can see that these measures also vary considerably across
subperiods. The picture is quite unambiguous with respect to stock portfolios. There is
excess positive kurtosis and negative skew. These indicate extreme gains and, even more so,
extreme losses that are significantly more likely than would be predicted by the normal dis-
tribution. We must conclude that VaR (and similar risk measures) to augment standard
deviation is in order.
The last section in Table 5.2 presents performance statistics, Sharpe ratios, and value at
risk. Sharpe ratios of stock portfolios are in the range of 0.37–0.39 for the overall history
and range between 0.34–0.46 across all subperiods. We can estimate that the return-risk
trade- off in stocks on an annual basis is about a .4% risk premium for each increment of 1%
to standard deviation. In fact, just as with the average excess return, the differences between
subperiods are not significant. The same can be said about the three stock portfolios: None
showed significant superior performance. Bonds can outperform stocks in periods of falling
interest rates, as we see from the Sharpe ratios in the most recent subperiod. But, as noted
earlier, bond portfolios are not sufficiently diversified to allow for the use of the Sharpe ratio
as a performance measure. (As we will discuss in later chapters, standard deviation as a risk
measure makes sense for an investor’s overall portfolio but not for one relatively narrow
com- ponent of it.)
The VaR panel in Table 5.2 shows unambiguously for stocks, and almost so for bonds, that
potential losses are larger than suggested by likewise normal distributions. To highlight this
observation, the last panel of the table shows the difference of actual 5% VaR from likewise
normal distributions; the evidence is quite clear and consistent with the kurtosis and skew
statistics.
Finally, investing internationally is no longer considered exotic, and Table 5.2 also
provides some information on the historical results from international investments. It
appears that for passive investors who focus on investments in index funds, international
diversification doesn’t deliver impressive improvement over investments in the U.S. alone.
However, international investments do hold large potential for active investors. We elaborate
on these observations in Chapter 19, which is devoted to international investing.
CONCEPT
c h e ck 5.4 Compute the average excess return on large-company stocks (over the T-bill rate) and the
standard deviation for the years 1926–1934. You will need to obtain data from the spread-
sheet available at the Online Learning Center at www.mhhe.com/bkm. Look for Chapter 5
material.
5.4 INFLATION AND REAL RATES OF RETURN

A 10% annual rate of return means that your investment was worth 10% more at the end of
the year than it was at the beginning of the year. This does not necessarily mean, however,
that you could have bought 10% more goods and services with that money, for it is possible
that in the course of the year prices of goods also increased. If prices have changed, the
increase in your purchasing power will not match the increase in your dollar wealth.
At any time, the prices of some goods may rise while the prices of other goods may fall;
the general trend in prices is measured by examining changes in the consumer price index, or
CPI. The CPI measures the cost of purchasing a representative bundle of goods, the
inflation rate “consumption basket” of a typical urban family of four. The inflation rate is measured by the
The rate at which prices are
rate of increase of the CPI. Suppose the rate of inflation (the percentage change in the CPI,
rising, measured as the rate denoted by i) for the last year amounted to i 5 6%. The purchasing power of money was thus
of increase of the CPI. reduced by 6%. Therefore, part of your investment earnings were offset by the reduction in
the purchasing power of the
dollars you received at the end of the year. With a 10% interest rate, for example, after you
netted out the 6% reduction in the purchasing power of money, you were left with a net
increase in purchasing power of about 4%. Thus, we need to distinguish between a nominal
nominal interest rate
interest rate—the growth rate of money—and a real interest rate—the growth rate of
The interest rate in terms of
purchasing power. If we call R the nominal rate, r the real rate, and i the inflation rate, then
nominal (not adjusted for
we conclude
purchasing power) dollars.
r<R2i (5.15)
real interest rate
The excess of the interest rate
In words, the real rate of interest is the nominal rate reduced by the loss of purchasing power
over the inflation rate. The
resulting from inflation.
growth rate of purchasing
In fact, the exact relationship between the real and nominal interest rates is given by
power derived from an
11R investment.
11r5 (5.16)
11i
In words, the growth factor of your purchasing power, 1 1 r, equals the growth factor of
your money, 1 1 R, divided by the new price level that is 1 1 i times its value in the
previous period. The exact relationship can be rearranged to
R2i
r5 11i (5.17)
which shows that the approximate rule overstates the real rate by the factor 1 1
i.9
If the interest rate on a one-year CD is 8%, and you expect inflation to be 5% over the coming EXAMPLE 5.6
year, then using the approximation given in Equation 5.15, you expect the real rate to be r 5 8% 2
5% 5 3%. Real versus
.08 2 .05 Nominal Rates
Using the exact formula given in Equation 5.17, the real rate is r 5 5 .0286 , or 2.86%.
1 1 .05
Therefore, the approximation rule overstates the expected real rate by only .14 percentage
points. The approximation rule of Equation 5.16 is more accurate for small inflation rates and is
perfectly exact for continuously compounded rates.
The Equilibrium Nominal Rate of Interest

We’ve seen that the real rate of return is approximately the nominal rate minus the inflation
rate. Because investors should be concerned with real returns—the increase in their purchas-
ing power—they will demand higher nominal rates of return on their investments. This
higher rate is necessary to maintain the expected real return as inflation increases.
Irving Fisher (1930) argued that the nominal rate ought to increase one-for-one with
increases in the expected inflation rate. Using E(i) to denote the current expected inflation
over the coming period, then the so-called Fisher equation is
R 5 r 1 E(i) (5.18)
Suppose the real rate of interest is 2%, and the inflation rate is 4%, so that the nominal
interest rate is about 6%. If the expected inflation rate rises to 5%, the nominal interest rate
should climb to roughly 7%. The increase in the nominal rate offsets the increase in
expected inflation, giving investors an unchanged growth of purchasing power at a 2% rate.
9
Notice that for continuously compounded rates, Equation 5.16 is perfectly accurate. Because ln(x/y) 5 ln(x) 2
ln(y), the continuously compounded real rate of return, rcc, can be derived from the annual rates as
11R
rcc 5 ln(1 1 r) 5 lna b 5 ln(1 1 R) 2 ln(1 1 i) 5 R c 2 icc
11i c
CONCEPT
c h e ck 5.5 a. Suppose the real interest rate is 3% per year, and the expected inflation rate is 8%.
What is the nominal interest rate?
b. Suppose the expected inflation rate rises to 10%, but the real rate is unchanged. What
happens to the nominal interest rate?
U.S. History of Interest Rates, Inflation,

and Real Interest Rates
Figure 5.5 plots nominal interest rates, inflation rates, and real rates in the U.S. between 1926
and 2010. Since the mid-1950s, nominal rates have increased roughly in tandem with
inflation, broadly consistent with the Fisher equation. The 1930s and 1940s, however, show
us that very volatile levels of unexpected inflation can play havoc with realized real rates of
return.
Table 5.3 quantifies what we see in Figure 5.5. One interesting pattern that emerges is the
steady increase in the average real interest rate across the three subperiods reported in the
table. Perhaps this reflects the shrinking national savings rate (and therefore reduced
availability of funds to borrowers) over this period. Another striking observation from Table
5.3 is the dramatic reduction in the variability of the inflation rate and the real interest rate.
This is reflected in the decline in standard deviations as well as in the steady attenuation of
minimum and maximum values. This reduction in variability also is related to the patterns in
correlation that we observe. According to the Fisher equation, an increase in expected
inflation translates directly into an increase in nominal interest rates; therefore, the
correlation between nominal rates and inflation rates should be positive and high. In
contrast, the correlation between real rates and inflation should be zero, because expected
inflation is fully factored into the nominal interest rate and does not affect the expected real
rate of return. The table indicates that during the early period, 1926–1955, market rates did
not accord to this logic, possibly due to the extraordinarily high and almost certainly
unforeseen variability in inflation rates. Since 1955, however, the nominal T-bill rate and
inflation rate have tracked each other far more closely (as is clear from Figure 5.5), and the
correlations show greater consistency with Fisher’s logic.
Inflation-indexed bonds called Treasury Inflation-Protected Securities (TIPS) were
introduced in the U.S. in 1997. These are bonds of 5- to 30-year original maturities with
coupons and principal that increase at the rate of inflation. (We discuss these bonds in more
detail in Chapter 10.) The difference between nominal rates on conventional T-bonds and
the rates on equal-maturity TIPS provides a measure of expected inflation (often called
break-even inflation) over that maturity.
FIGURE 5.5
20
Interest rates, inflation, T-bills
and real interest Inflation
rates, 1926–2010 15
Real T-bills
Source: T-bills: Prof. Kenneth
10
French, http://mba.tuck.dart
mouth.edu/pages/faculty/ken
Percent
.french/data_library.html; 5
Inflation: Bureau of Labor
Statistics, www.bls.gov; Real 0
rate: authors’ calculations.
25
210
215
220
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
TABLE 5.3 Annual rates of return statistics for U.S. T-bills, inflation,
and real interest rates, 1926–2010 and three subperiods
(%)
U.S. Market
T-Bills Inflation Real T-Bills
Arithmetic average
1926–2010 3.66 3.08 0.68
1926–1955 1.10 1.51 20.11
1956–1985 5.84 4.85 0.98
1986–2010 4.14 2.84 1.26
Standard deviation
1926–2010 3.09 4.17 3.89
1926–1955 1.22 5.55 5.84
1956–1985 3.19 3.50 2.39
1986–2010 2.25 1.31 1.87
Correlations T-bills1inflation Real bills1inflation
1926–2010 0.41 20.46
1926–1955 20.30 20.59
1956–1985 0.72 20.53
1986–2010 0.53 0.35
Minimum (lowest rate)
1926–2010 20.04 210.27 215.04
1926–1955 20.04 210.27 215.04
1956–1985 1.53 0.67 23.65
1986–2010 0.05 20.04 22.64
Maximum (highest rate)
1926–2010 14.72 18.13 12.50
1926–1955 4.74 18.13 12.50
1956–1985 14.72 13.26 6.45
1986–2010 8.38 6.26 4.91
* Two slightly negative interest rates occurred in the 1930s, before T-bills were introduced. In those days, the Treasury instead
guaranteed short-term bonds. In highly uncertain times, great demand for these bonds could result in a negative rate.
Source: T-bills: Fama and French risk-free rate; Inflation data: Bureau of Labor Statistics (inflation-cpiu-dec2dec).
ASSET ALLOCATION ACROSS RISKY

5.5
AND RISK-FREE PORTFOLIOS
History shows us that long-term bonds have been riskier investments than investments in
Treasury bills and that stock investments have been riskier still. On the other hand, the
riskier investments have offered higher average returns. Investors, of course, do not make
all-or- nothing choices from these investment classes. They can and do construct their
portfolios using securities from all asset classes.
A simple strategy to control portfolio risk is to specify the fraction of the portfolio
invested in broad asset classes such as stocks, bonds, and safe assets such as Treasury bills.
This aspect of portfolio management is called asset allocation and plays an important role
in the determination of portfolio performance. Consider this statement by John Bogle, made asset allocation
when he was the chairman of the Vanguard Group of Investment Companies: Portfolio choice among broad
investment classes.
The most fundamental decision of investing is the allocation of your assets: How much should
you own in stock? How much should you own in bonds? How much should you own in cash
reserves? . . . That decision [has been shown to account] for an astonishing 94% of the
differences
in total returns achieved by institutionally managed pension funds.........There is no reason to

believe that the same relationship does not also hold true for individual investors.10
The most basic form of asset allocation envisions the portfolio as dichotomized into risky
capital allocation versus risk-free assets. The fraction of the portfolio placed in risky assets is called the
The choice between risky capital allocation to risky assets and speaks directly to investor risk aversion.
and risk-free assets. To focus on the capital allocation decision, we think about an investor who allocates
funds between T-bills and a portfolio of risky assets. We can envision the risky portfolio, P,
as a mutual fund or ETF (exchange-traded fund) that includes a bundle of risky assets in
desired, fixed proportions. Thus, when we shift wealth into and out of P, we do not change
the relative proportion of the various securities within the risky portfolio. We put off until
the next chapter the question of how to best construct the risky portfolio. We call the overall
complete portfolio portfolio composed of the risk-free asset and the risky portfolio, P, the complete portfolio
The entire portfolio including that includes the entire investor’s wealth.
risky and risk-free assets.
The Risk-Free Asset
The power to tax and to control the money supply lets the government, and only the
government, issue default-free (Treasury) bonds. The default-free guarantee by itself is not
sufficient to make the bonds risk-free in real terms, since inflation affects the purchasing power
of the proceeds from the bonds. The only risk-free asset in real terms would be a price-indexed
government bond such as TIPS. Even then, a default-free, perfectly indexed bond offers a
guaranteed real rate to an investor only if the maturity of the bond is identical to the investor’s
desired holding period. These qualifications notwithstanding, it is common to view Treasury
bills as the risk-free asset. Any inflation uncertainty over the course of a few weeks, or even
months, is negligible compared to the uncertainty of stock market returns.11
In practice, most investors treat a broader range of money market instruments as
effectively risk-free assets. All the money market instruments are virtually immune to
interest rate risk (unexpected fluctuations in the price of a bond due to changes in market
interest rates) because of their short maturities, and all are fairly safe in terms of default or
credit risk.
Money market mutual funds hold, for the most part, three types of securities: Treasury
bills, bank certificates of deposit (CDs), and commercial paper. The instruments differ
slightly in their default risk. The yields to maturity on CDs and commercial paper, for
identical maturities, are always slightly higher than those of T-bills. A history of this yield
spread for 90-day CDs is shown in Figure 2.2 in Chapter 2.
Money market funds have changed their relative holdings of these securities over time,
but by and large, the risk of such blue-chip, short-term investments as CDs and commercial
paper is minuscule compared to that of most other assets, such as long-term corporate bonds,
common stocks, or real estate. Hence, we treat money market funds, as well as T-bills, as
representing the most easily accessible risk-free asset for most investors.
Portfolio Expected Return and Risk

We can examine the risk-return combinations that result from various capital allocations in
the complete portfolio to risky versus risk-free assets. Finding the available combinations of
10
John C. Bogle, Bogle on Mutual Funds (Burr Ridge, IL: Irwin Professional Publishing, 1994), p. 235.
11
In the wake of the euro crisis as well as the credit downgrade of the United States in the summer of 2011, one
clearly needs to consider whether (or when) sovereign debt can be treated as risk-free. Governments that issue debt
in their home currency can in principle always repay that debt, if need be by printing more money in that currency.
This strategy, however, can lead to runaway inflation, so the real return on that debt would hardly be risk-free.
More- over, the cost of possible hyperinflation can be so great that they might justifiably conclude that default is the
lesser of the two evils. Governments that issue debt in currencies they do not control (e.g., euro-denominated Greek
debt) cannot fall back on the printing press, even under extreme duress, so default in that situation is certainly
possible. Since the euro crisis, analysts have focused considerable attention on measures of sovereign fiscal health
such as the ratio of indebtedness to GDP. As is also true of corporate debt, long- and medium-term debt issues are
typically riskier, as they allow more time for credit conditions to deteriorate before the loan is paid off.
risk and return is the “technical” part of capital allocation; it deals only with the
opportunities available to investors. In the next section, we address the “personal
preference” part of the problem, the individual’s choice of the preferred risk-return
combination, given his degree of risk aversion.
Since we assume that the composition of the risky portfolio, P, already has been
determined, the only concern here is with the proportion of the investment budget (y) to
be allocated to it. The remaining proportion (1 2 y) is to be invested in the risk-free
asset, which has a rate of return denoted rf .
We denote the actual risky rate of return by rP, the expected rate of return on P by E(rP),
and its standard deviation by sP. In the numerical example, E(rP) 5 15%, sP 5 22%, and rf 5
7%. Thus, the risk premium on the risky asset is E(rP) 2 rf 5 8%.
Let’s start with two extreme cases. If you invest all of your funds in the risky asset, that is,
if you choose y 5 1, the expected return on your complete portfolio will be 15% and the
standard deviation will be 22%. This combination of risk and return is plotted as point P
in Figure 5.6. At the other extreme, you might put all of your funds into the risk-free
asset, that is, you choose y 5 0. In this case, you would earn a riskless return of 7%.
(This choice is plotted as point F in Figure 5.6.)
Now consider more moderate choices. For example, if you allocate equal amounts of your
complete portfolio, C, to the risky and risk-free assets, that is, you choose y 5 .5, the
expected return on the complete portfolio will be the average of E(rP) and rf .Therefore,
E(rC) 5 .5 3 7% 1
.5 3 15% 5 11%. The risk premium of the complete portfolio is therefore 11% 2 7% 5
4%, which is half of the risk premium of P. The standard deviation of the portfolio also
is one-half of P ’s, that is, 11%. When you reduce the fraction of the complete portfolio
allocated to the risky asset by half, you reduce both the risk and risk premium by half.
To generalize, the risk premium of the complete portfolio, C, will equal the risk premium
of the risky asset times the fraction of the portfolio invested in the risky asset.
E(rC) 2 rf 5 y[E(rP) 2 rf ] (5.19)
The standard deviation of the complete portfolio will equal the standard deviation of the
risky asset times the fraction of the portfolio invested in the risky asset.
sC 5 y sP (5.20)
In sum, both the risk premium and the standard deviation of the complete portfolio increase
in proportion to the investment in the risky portfolio. Therefore, the points that describe the
risk and return of the complete portfolio for various capital allocations of y all plot on the
FIGURE 5.6
E(r) The investment
opportunity set with a
risky asset and a risk-free
asset
CAL = Capital
allocation
P line
E(rP ) = 15% y = 1.25
y = .50
E(rP) – rƒ =
S = 8/22 8%
rƒ = 7%
F
σ
σP = 22%
straight line connecting F and P, as shown in Figure 5.6, with an intercept of rf and slope
(rise/run) equal to the familiar Sharpe ratio of P :
E(rP) 2 rf 15 2 7
S5 5 5 .36 (5.21)
sP 22
CONCEPT
c h e ck 5.6 What are the expected return, risk premium, standard deviation, and ratio of risk premium
to standard deviation for a complete portfolio with y 5 .75?
The Capital Allocation Line

The line plotted in Figure 5.6 depicts the risk-return combinations available by varying
capital allocation, that is, by choosing different values of y. For this reason it is called the
capital allocation line capital allocation line, or CAL. The slope, S, of the CAL equals the increase in expected
(CAL) return that an investor can obtain per unit of additional standard deviation or extra return
Plot of risk-return per extra risk. It is obvious why it is also called the reward-to-volatility ratio, or Sharpe
combinations available by ratio, after William Sharpe who first suggested its use.
varying portfolio allocation Notice that the Sharpe ratio is the same for risky portfolio P and the complete portfolio
between a risk-free asset that mixes P and the risk-free asset in equal proportions.
and a risky portfolio.
Expected
Return Risk Standard Reward-to-
Premium Deviation Volatility Ratio
Portfolio P: 15% 8% 22% 8

5 0.36
22
Portfolio C: 11% 4% 11% 4
5 0.36
11
In fact, the reward-to-volatility ratio is the same for all complete portfolios that plot on the
capital allocation line. While the risk-return combinations differ according to the investor’s
choice of y, the ratio of reward to risk is constant.
What about points on the CAL to the right of portfolio P in the investment
opportunity set? You can construct complete portfolios to the right of point P by
borrowing, that is, by choosing y . 1. This means that you borrow a proportion of y
2 1 and invest both the bor- rowed funds and your own wealth in the risky portfolio P.
If you can borrow at the risk-free rate, rf 5 7%, then your rate of return will be rC 5 2(
y 2 1)rf 1 y rP 5 rf 1 y(rP 2 rf ). This complete portfolio has risk premium of y[E(rP) 2 rf
] and SD 5 y sP. Verify that your Sharpe ratio equals that of any other portfolio on the
same CAL.
EXAMPLE 5.7
Suppose the investment budget is $300,000, and an investor borrows an additional $120,000,
Levered Complete investing the $420,000 in the risky asset. This is a levered position in the risky asset, which is
financed in part by borrowing. In that case
Portfolios
420,000
y 5 300,000 5 1.4
and 1 2 y 5 1 2 1.4 5 2.4, reflecting a short position in the risk-free asset, or a borrowing position.
Rather than lending at a 7% interest rate, the investor borrows at 7%. The portfolio rate of return
is
E( rC ) 5 7 1 (1.4 3 8) 5 18.2
Another way to find this portfolio rate of return is as follows: You expect to earn $63,000 (15% of
$420,000) and pay $8,400 (7% of $120,000) in interest on the loan. Simple subtraction yields an
(continued)
expected profit of $54,600, which is 18.2% of your investment budget of $300,000. Therefore, EXAMPLE 5.7
E(rC) 5 18.2%.
Your portfolio still exhibits the same reward-to-volatility ratio: Levered Complete
Portfolios
s C5 1.4 3 22 5 30.8
(concluded)
E( r ) 2 r 11.2
C f
S5 sC 5 5 .36
30.8
As you might have expected, the levered portfolio has both a higher expected return and a
higher standard deviation than an unlevered position in the risky asset.
Risk Aversion and Capital Allocation

We have developed the CAL, the graph of all feasible risk-return combinations available
from allocating the complete portfolio between a risky portfolio and a risk-free asset. The
investor confronting the CAL now must choose one optimal combination from the set
of feasible choices. This choice entails a trade-off between risk and return. Individual
investors with different levels of risk aversion, given an identical capital allocation line, will
choose different positions in the risky asset. Specifically, the more risk-averse investors
will choose to hold less of the risky asset and more of the risk-free asset.
How can we find the best allocation between the risky portfolio and risk-free asset?
Recall that a particular investor’s degree of risk aversion (A) measures the price of risk she
demands from the complete portfolio in which her entire wealth is invested. The
compensation for risk demanded by the investor must be compared to the price of risk
offered by the risky portfolio,
P. We can find the investor’s preferred capital allocation, y, by dividing the risky
portfolio’s price of risk by the investor’s risk aversion, her required price of risk:
Available risk premium to variance ratio

[E(rP) 2 rf [E(rP) 2 rf ]
y 5 Required risk premium to variance ratio 5 ]/sP2 5 (5.22)
As2P
A
Notice that when the price of risk of the available risky portfolio exactly matches the
investor’s degree of risk aversion, her entire wealth will be invested in it (y 5 1).
What would the investor of Equation 5.12 (with A 5 3.91) do when faced with the
market index portfolio of Equation 5.13 (with price of risk 5 2)? Equation 5.22 tells us
that this investor would invest y 5 2/3.91 5 0.51 (51%) in the market index portfolio and
a proportion 1 2 y 5 0.49 in the risk-free asset.
Graphically, more risk-averse investors will choose portfolios near point F on the capital
allocation line plotted in Figure 5.6. More risk-tolerant investors will choose points closer to
P, with higher expected return and higher risk. The most risk-tolerant investors will choose
portfolios to the right of point P. These levered portfolios provide even higher expected
returns, but even greater risk.
The investor’s asset allocation choice also will depend on the trade-off between risk and
return. When the reward-to-volatility ratio increases, investors might well decide to take on
riskier positions. Suppose an investor reevaluates the probability distribution of the risky
portfolio and now perceives a greater expected return without an accompanying increase in
the standard deviation. This amounts to an increase in the reward-to-volatility ratio or,
equivalently, an increase in the slope of the CAL. As a result, this investor will choose a
higher y, that is, a greater position in the risky portfolio.
One role of a professional financial adviser is to present investment opportunity alterna-
tives to clients, obtain an assessment of the client’s risk tolerance, and help determine the
appropriate complete portfolio.12
12
“Risk tolerance” is simply the flip side of “risk aversion.” Either term is a reasonable way to describe attitudes
toward risk. We generally find it easier to talk about risk aversion, but practitioners often use the term risk
tolerance.
5.6 PASSIVE STRATEGIES AND THE CAPITAL MARKET LINE

passive strategy A passive strategy is based on the premise that securities are fairly priced, and it avoids the
Investment policy that costs involved in undertaking security analysis. Such a strategy might at first blush appear to
avoids security analysis. be naive. However, we will see in Chapter 8 that intense competition among professional
money managers might indeed force security prices to levels at which further security
analysis is unlikely to turn up significant profit opportunities. Passive investment strategies
may make sense for many investors.
To avoid the costs of acquiring information on any individual stock or group of stocks,
we may follow a “neutral” diversification approach. Select a diversified portfolio of
common stocks that mirrors the corporate sector of the broad economy. This results in a
value-weighted portfolio, which, for example, invests a proportion in GE stock that equals
the ratio of GE’s market value to the market value of all listed stocks.
Such strategies are called indexing. The investor chooses a portfolio of all the stocks in a
broad market index such as the S&P 500. The rate of return on the portfolio then replicates
the return on the index. Indexing has become a popular strategy for passive investors. We
call the capital allocation line provided by one-month T-bills and a broad index of common
capital market line stocks the capital market line (CML). That is, a passive strategy using the broad stock
The capital allocation line market index as the risky portfolio generates an investment opportunity set that is
using the market index represented by the CML.
portfolio as the risky asset.
Historical Evidence on the Capital Market Line
Table 5.4 is a small cut-and-paste from Table 5.3, which concentrates on S&P 500 data, a
popular choice for a broad stock-market index. As we discussed earlier, the large standard
deviation of its rate of return implies that we cannot reject the hypothesis that the entire 85-
year period is characterized by the same Sharpe ratio. Using this history as a guide, inves-
tors might reasonably forecast a risk premium of around 8% coupled with a standard
deviation of approximately 20%, resulting in a Sharpe ratio of .4.
We also have seen that to hold a complete portfolio with these risk-return characteristics,
the “average” investor (with y 5 1) would need to have a coefficient of risk aversion of
.08/.202 5 2. But that average investor would need some courage. As the VaR figures in Table
5.4 indicate, the market index has exhibited a probability of 5% of a 36.86% or worse loss in
a year; surely this is no picnic. This substantial risk, together with differences in risk
aversion across individuals, might explain the large differences we observe in portfolio
positions across investors.
Finally, notice the instability of the excess returns on the S&P 500 across the 30-year sub-
periods in Table 5.4. The great variability in excess returns raises the question of whether the
8% historical average really is a reasonable estimate of the risk premium looking into the
future. It also suggests that different investors may come to different conclusions about
future excess returns, another reason for capital allocations to vary.
In fact, there has been considerable recent debate among financial economists about the
“true” equity risk premium, with an emerging consensus that the historical average may be
an unrealistically high estimate of the future risk premium. This argument is based on
several
TABLE 5.4 Excess return statistics for the S&P 500
Excess Return (%)
Average Std Dev Sharpe Ratio 5% VaR
1926–2010 8.00 20.70 .39 236.86

1926–1955 11.67 25.40 .46 253.43
1956–1985 5.01 17.58 .28 230.51
1986–2010 7.19 17.83 .40 242.28
On the MARKET FRONT
TRIUMPH OF THE OPTIMISTS On the other hand, the equity risk premium is probably not as
As a whole, the last eight decades have been very kind to U.S. large as the post-1926 evidence from Table 5.3 would seem to
equity investors. Even accounting for miserable 2008 returns, indicate. First, results from the first 25 years of the last century
stock investments have outperformed investments in safe (which included the first World War) were less favorable to stocks.
Treasury bills by more than 7% per year. The real rate of return Second,
averaged more than 6%, implying an expected doubling of the U.S. returns have been better than those of most other countries,
real value of the investment portfolio about every 12 years! and so a more representative value for the historical risk premium
Is this experience representative? A book by three professors may be lower than the U.S. experience. Finally, the sample that is
at the London Business School, Elroy Dimson, Paul Marsh, and amenable to historical analysis suffers from a self-selection
Mike Staunton, extends the U.S. evidence to other countries and problem. Only those markets that have survived to be studied can
to longer time periods. Their conclusion is given in the book’s title, be included in the analysis. This leaves out countries such as
Triumph of the Optimists*: In every country in their study (which Russia or China, whose markets were shut down during
included markets in North America, Europe, Asia, and Africa), communist rule, and whose results if included would surely bring
the investment optimists—those who bet on the economy by down the average historical performance of equity investments.
investing in stocks rather than bonds or bills—were vindicated. Nevertheless, there is powerful evidence of a risk premium that
Over the long haul, stocks beat bonds everywhere. shows its force everywhere the authors looked.
*Elroy Dimson, Paul Marsh, Mike Staunton, Triumph of the Optimists: 101
Years of Global Investment Returns (Princeton, NJ: Princeton University Press,
2002).
factors: the use of longer time periods in which equity returns are examined; a broad range mana
of countries rather than just the U.S. in which excess returns are computed (Dimson, Marsh, geme
and Staunton, 2001); direct surveys of financial executives about their expectations for stock nt
market returns (Graham and Harvey, 2001); and inferences from stock market data about indus
investor expectations ( Jagannathan, McGrattan, and Scherbina, 2000; Fama and French, try
2002). The nearby box discusses some of this evidence. goin
g are
(1)
Costs and Benefits of Passive Investing
the
The fact that an individual’s capital allocation decision is hard does not imply that its imple- large
mentation needs to be complex. A passive strategy is simple and inexpensive to implement: pote
Choose a broad index fund or ETF and divide your savings between it and a money market ntial
fund. To justify spending your own time and effort or paying a professional to pursue an of
active strategy requires some evidence that those activities are likely to be profitable. As we enric
shall see later in the text, this is much harder to come by than you might expect! hme
To choose an active strategy, an investor must be convinced that the benefits outweigh nt
the cost, and the cost can be quite large. As a benchmark, annual expense ratios for index from
funds are around 20 and 50 basis points for U.S. and international stocks, respectively. The succ
cost of utiliz- ing a money market fund is smaller still, and T-bills can be purchased at no essfu
cost. l
Here is a very cursory idea of the cost of active strategies: The annual expense ratio of an inves
active stock mutual fund averages around 1% of invested assets, and mutual funds that invest in tmen
more exotic assets such as real estate or precious metals can be more expensive still. A hedge ts—
fund will cost you 1% to 2% of invested assets plus 10% or more of any returns above the the
risk-free rate. If you are wealthy and seek more dedicated portfolio management, costs will be same
even higher. powe
Because of the power of compounding, an extra 1% of annual costs can have large conse- r of
quences for the future value of your portfolio. With a risk-free rate of 2% and a risk com
premium of 8%, you might expect your wealth to grow by a factor of 1.10 30 5 17.45 over a poun
30-year investment horizon. If fees are 1%, then your net return is reduced to 9%, and your ding
wealth grows by a factor of only 1.0930 5 13.26 over that same horizon. That seemingly work
small management fee reduces your final wealth by about one-quarter. s in
The potential benefits of active strategies are discussed in detail in Chapter 8. The news your
is generally not that good for active investors. However, the factors that keep the active favor
if
you can add even a
139
few basis points to total return, (2) the difficulty in assessing performance (discussed in
Chapter 18), and (3) uninformed investors who are willing to pay for professional money
management. There is no question that some money managers can outperform passive
strategies. The problem is (1) how do you identify them and (2) do their fees outstrip their
potential. Whatever the choice one makes, one thing is clear: The CML using the passive
market index is not an obviously inferior choice.
SUMMARY • Investors face a trade-off between risk and expected return. Historical data confirm our
intuition that assets with low degrees of risk should provide lower returns on average
than do those of higher risk.
• Shifting funds from the risky portfolio to the risk-free asset is the simplest way to reduce
risk. Another method involves diversification of the risky portfolio. We take up
diversification in later chapters.
• U.S. T-bills provide a perfectly risk-free asset in nominal terms only. Nevertheless, the
standard deviation of real rates on short-term T-bills is small compared to that of assets
such as long-term bonds and common stocks, so for the purpose of our analysis, we con-
sider T-bills the risk-free asset. Besides T-bills, money market funds hold short-term,
safe obligations such as commercial paper and CDs. These entail some default risk but
relatively little compared to most other risky assets. For convenience, we often refer to
money market funds as risk-free assets.
• A risky investment portfolio (referred to here as the risky asset) can be characterized by
its reward-to-volatility ratio. This ratio is the slope of the capital allocation line (CAL),
the line connecting the risk-free asset to the risky asset. All combinations of the risky and
risk-free asset lie on this line. Investors would prefer a steeper-sloping CAL, because that
means higher expected returns for any level of risk.
• An investor’s preferred choice among the portfolios on the capital allocation line will
depend on risk aversion. Risk-averse investors will weight their complete portfolios
more heavily toward Treasury bills. Risk-tolerant investors will hold higher proportions
of their complete portfolios in the risky asset.
• The capital market line is the capital allocation line that results from using a passive
investment strategy that treats a market index portfolio, such as the Standard & Poor’s
500, as the risky asset. Passive strategies are low-cost ways of obtaining well-
diversified portfolios with performance that will reflect that of the broad stock market.
KEY TERMS arithmetic average, geometric average, 112 real interest rate, 131
112 asset allocation,
holding-period return risk aversion, 122
133
(HPR), 111 risk-free rate, 122
capital allocation, 134
inflation rate, 130 risk premium, 122
capital allocation line
kurtosis, 121 scenario analysis, 115
(CAL), 136
mean-variance analysis, 125 Sharpe (or reward-to-
capital market line, 138
nominal interest rate, 131 volatility) ratio, 125
complete portfolio, 134
passive strategy, 138 skew, 121
dollar-weighted average
price of risk, 123 standard deviation, 116
return, 113
www.mhhe.com/bkm
probability value at risk (VaR),

excess return, 122
distribution, 115 118 variance, 116
expected return, 115
KEY FORMULAS Arithmetic average of n returns: (r1 1 r2 1 # # # 1 rn)/n

Geometric average of n returns: [(1 1 r1) (1 1 r2) # # # (1 1 rn)]1/n 2 1
Continuously compounded rate of return, rcc: ln(1 1 Effective annual
rate) Expected return: S [prob(Scenario) 3 Return in scenario]
Variance: S [prob(Scenario) 3 (Deviation from mean in
scenario)2] Standard deviation: "Variance
Portfolio risk premium E(rP) 2 rf

Sharpe ratio: Standard deviation of excess return
5 sP
1 1 Nominal return
Real rate of return: 21
1 1 Inflation
rate
Real rate of return (continuous compounding): rnominal 2 Inflation rate
E(r ) 2 r
P f
Optimal capital allocation to the risky asset, y: A
2 P
s
Select problems are available in McGraw-Hill’s

Connect Finance. Please see the Supplements PROBLEM SETS
section of the book’s frontmatter for more information.
Basic
1. Suppose you’ve estimated that the fifth-percentile value at risk of a portfolio is 230%.
Now you wish to estimate the portfolio’s first-percentile VaR (the value below which lie
1% of the returns). Will the 1% VaR be greater or less than 230%? (LO 5-2)
2. To estimate the Sharpe ratio of a portfolio from a history of asset returns, we use the
difference between the simple (arithmetic) average rate of return and the T-bill rate.
Why not use the geometric average? (LO 5-4)
3. When estimating a Sharpe ratio, would it make sense to use the average excess real
return that accounts for inflation? (LO 5-4)
4. You’ve just decided upon your capital allocation for the next year, when you realize that
you’ve underestimated both the expected return and the standard deviation of your risky
portfolio by 4%. Will you increase, decrease, or leave unchanged your allocation to risk-
free T-bills? (LO 5-4)
Intermediate
5. Suppose your expectations regarding the stock market are as follows:
State of the Economy Probability HPR
Boom 0.3 44%

Normal growth 0.4 14
Recession 0.3 216
Use Equations 5.6–5.8 to compute the mean and standard deviation of the HPR on
stocks. (LO 5-4)
6. The stock of Business Adventures sells for $40 a share. Its likely dividend payout
and end-of-year price depend on the state of the economy by the end of the year as
www.mhhe.com/bkm
follows: (LO 5-2)
Dividend Stock Price
Boom $2.00 $50

Normal economy 1.00 43
Recession .50 34
a. Calculate the expected holding-period return and standard deviation of the holding-
period return. All three scenarios are equally likely.
142 Part TWO Portfolio Theory
b. Calculate the expected return and standard deviation of a portfolio invested half in
Business Adventures and half in Treasury bills. The return on bills is 4%.
7. XYZ stock price and dividend history are as follows:
Year Beginning-of-Year Price Dividend Paid at Year-End
2010 $100 $4
2011 $110 $4
2012 $ 90 $4
2013 $ 95 $4
An investor buys three shares of XYZ at the beginning of 2010, buys another two shares
at the beginning of 2011, sells one share at the beginning of 2012, and sells all four
remaining shares at the beginning of 2013. (LO 5-1)
a. What are the arithmetic and geometric average time-weighted rates of return for the
investor?
b. What is the dollar-weighted rate of return? (Hint: Carefully prepare a chart of cash
flows for the four dates corresponding to the turns of the year for January 1, 2010,
to January 1, 2013. If your calculator cannot calculate internal rate of return, you
will have to use a spreadsheet or trial and error.)
8. a. Suppose you forecast that the standard deviation of the market return will be
20% in the coming year. If the measure of risk aversion in Equation 5.13 is A 5
4, what would be a reasonable guess for the expected market risk premium?
b. What value of A is consistent with a risk premium of 9%?
c. What will happen to the risk premium if investors become more risk tolerant?
(LO 5-4)
9. Using the historical risk premiums as your guide, what is your estimate of the
expected annual HPR on the S&P 500 stock portfolio if the current risk-free interest
rate is 5%? (LO 5-3)
10. What has been the historical average real rate of return on stocks, Treasury bonds,
and Treasury bills? (LO 5-2)
11. Consider a risky portfolio. The end-of-year cash flow derived from the portfolio will
be either $50,000 or $150,000, with equal probabilities of .5. The alternative riskless
investment in T-bills pays 5%. (LO 5-3)
a. If you require a risk premium of 10%, how much will you be willing to pay for
the portfolio?
b. Suppose the portfolio can be purchased for the amount you found in (a). What will
the expected rate of return on the portfolio be?
c. Now suppose you require a risk premium of 15%. What is the price you will be
willing to pay now?
d. Comparing your answers to (a) and (c), what do you conclude about the relationship
between the required risk premium on a portfolio and the price at which the portfolio
will sell?
For Problems 12–16, assume that you manage a risky portfolio with an expected
rate of return of 17% and a standard deviation of 27%. The T-bill rate is 7%.
www.mhhe.com/bkm
12. Your client chooses to invest 70% of a portfolio in your fund and 30% in a T-bill
money market fund. (LO 5-3)
a. What is the expected return and standard deviation of your client’s portfolio?
b. Suppose your risky portfolio includes the following investments in the given proportions:
Stock A 27%
Stock B 33%
Stock C 40%
What are the investment proportions of your client’s overall portfolio, including the
position in T-bills?
c. What is the reward-to-volatility ratio (S) of your risky portfolio and your client’s
overall portfolio?
d. Draw the CAL of your portfolio on an expected return/standard deviation
diagram. What is the slope of the CAL? Show the position of your client on
your fund’s CAL.
13. Suppose the same client in the previous problem decides to invest in your risky
portfolio a proportion (y) of his total investment budget so that his overall portfolio will
have an expected rate of return of 15%. (LO 5-3)
a. What is the proportion y?
b. What are your client’s investment proportions in your three stocks and the T-bill
fund?
c. What is the standard deviation of the rate of return on your client’s portfolio?
14. Suppose the same client as in the previous problem prefers to invest in your portfolio a
proportion (y) that maximizes the expected return on the overall portfolio subject to the
constraint that the overall portfolio’s standard deviation will not exceed 20%. (LO 5-
3)
a. What is the investment proportion, y?
b. What is the expected rate of return on the overall portfolio?
15. You estimate that a passive portfolio invested to mimic the S&P 500 stock index yields
an expected rate of return of 13% with a standard deviation of 25%. Draw the CML and
your fund’s CAL on an expected return/standard deviation diagram. (LO 5-4)
a. What is the slope of the CML?
b. Characterize in one short paragraph the advantage of your fund over the
passive fund.
16. Your client (see previous problem) wonders whether to switch the 70% that is
invested in your fund to the passive portfolio. (LO 5-4)
a. Explain to your client the disadvantage of the switch.
b. Show your client the maximum fee you could charge (as a percent of the investment
in your fund deducted at the end of the year) that would still leave him at least as
well off investing in your fund as in the passive one. (Hint: The fee will lower the
slope of your client’s CAL by reducing the expected return net of the fee.)
17. What do you think would happen to the expected return on stocks if investors perceived
an increase in the volatility of stocks? (LO 5-4)
18. You manage an equity fund with an expected risk premium of 10% and a standard
deviation of 14%. The rate on Treasury bills is 6%. Your client chooses to invest
$60,000 of her portfolio in your equity fund and $40,000 in a T-bill money market
fund. What is the expected return and standard deviation of return on your client’s
portfolio? (LO 5-3)
19. What is the reward-to-volatility ratio for the equity fund in the previous problem?
(LO 5-4)
For Problems 20–22, download the spreadsheet containing the data for Table 5.2, “Rates
of return, 1926–2010,” from www.mhhe.com/bkm.
20. Calculate the same subperiod means and standard deviations for small stocks as Table
www.mhhe.com/bkm
Please visit us at
5.4 of the text provides for large stocks. (LO 5-2) www.mhhe.com/bkm
a. Have small stocks provided better reward-to-volatility ratios than large stocks?
b. Do small stocks show a similar higher standard deviation in the earliest subperiod
as Table 5.4 documents for large stocks?
Please visit us at
21. Convert the nominal returns on both large and small stocks to real rates. Reproduce www.mhhe.com/bkm
Table 5.4 using real rates instead of excess returns. Compare the results to those of
Table 5.4. (LO 5-1)
22. Repeat the previous problem for small stocks and compare with the results for Please visit us at
nominal rates. (LO 5-1) www.mhhe.com/bkm
Challenge
23. Download the annual returns on the combined NYSE/NASDAQ/AMEX markets as
well as the S&P 500 from the Online Learning Center at www.mhhe.com/bkm. For
Please visit us at
both indexes, calculate: (LO 5-2)
www.mhhe.com/bkm
a. Average return.
b. Standard deviation of return.
c. Skew of return.
d. Kurtosis of return.
e. The 5% value at risk.
f. Based on your answers to parts (b)–(e), compare the risk of the two indexes.
CFA Problems
1. A portfolio of nondividend-paying stocks earned a geometric mean return of 5%
between January 1, 2005, and December 31, 2011. The arithmetic mean return for
the same period was 6%. If the market value of the portfolio at the beginning of
2005 was $100,000, what was the market value of the portfolio at the end of
2011? (LO 5-1)
2. Which of the following statements about the standard deviation is/are true? A standard
deviation: (LO 5-2)
a. Is the square root of the variance.
b. Is denominated in the same units as the original data.
c. Can be a positive or a negative number.
3. Which of the following statements reflects the importance of the asset allocation
decision to the investment process? The asset allocation decision: (LO 5-3)
a. Helps the investor decide on realistic investment goals.
b. Identifies the specific securities to include in a portfolio.
c. Determines most of the portfolio’s returns and volatility over time.
d. Creates a standard by which to establish an appropriate investment time horizon.
Use the following data in answering CFA Questions 4–6.
Investment Expected E(r) Standard Deviation, s

Return,
1 .12 .30
2 .15 .50
3 .21 .16
4 .24 .21
Investor “satisfaction” with portfolio increases with expected return and decreases with
variance according to the “utility” formula: U 5 E(r) 2 ½ As2 where A 5 4.
4. Based on the formula for investor satisfaction or “utility,” which investment
would you select if you were risk averse with A 5 4? (LO 5-4)
5. Based on the formula above, which investment would you select if you were
www.mhhe.com/bkm
risk neutral? (LO 5-4)

6. The variable (A) in the utility formula represents the: (LO 5-4)
a. Investor’s return requirement.
b. Investor’s aversion to risk.
c. Certainty equivalent rate of the portfolio.
d. Preference for one unit of return per four units of risk.
Use the following scenario analysis for stocks X and Y to answer CFA
Questions 7 through 9.
Bear Market Normal Market Bull Market
Probability .2 .5 .3
Stock X 220% 18% 50%
Stock Y 215% 20% 10%
7. What are the expected returns for stocks X and Y ? (LO 5-2)
8. What are the standard deviations of returns on stocks X and Y ? (LO 5-2)
9. Assume that of your $10,000 portfolio, you invest $9,000 in stock X and $1,000
in stock Y. What is the expected return on your portfolio? (LO 5-3)
10. Probabilities for three states of the economy and probabilities for the returns on a
particular stock in each state are shown in the table below.
Probability of Stock
Probability of Stock Performance in Given
State of Economy Economic State Performance Economic State
Good .3 Good .6
Neutral .3
Poor .1
Neutral .5 Good .4
Neutral .3
Poor .3
Poor .2 Good .2
Neutral .3
Poor .5
What is the probability that the economy will be neutral and the stock will experience
poor performance? (LO 5-2)
11. An analyst estimates that a stock has the following probabilities of return depending on
the state of the economy. What is the expected return of the stock? (LO 5-2)
State of Economy Probability Return
Good .1 15%
Normal .6 13
Poor .3 7
WEB master
1. Use data from finance.yahoo.com to answer the following questions.
a. Select the Company tab and enter the ticker symbol “ADBE.” Click on the Profile tab
to see an overview of the company.
b. What is the latest price reported in the Summary section? What is the 12-month target
price? Calculate the expected holding-period return based on these prices.
www.mhhe.com/bkm
c. Use the Historical Prices section to answer the question “How much would I have
today if I invested $10,000 in ADBE five years ago?” Using this information,
calculate the five-year holding-period return on Adobe’s stock.
2. From the Historical Prices tab, download Adobe’s dividend-adjusted stock price for the
last 24 months into an Excel spreadsheet. Calculate the monthly rate of return for each
month, the average return, and the standard deviation of returns over that period.
3. Calculating the real rate of return is an important part of evaluating an investment’s
performance. To do this, you need to know the nominal return on your investment and
the rate of inflation during the corresponding period. To estimate the expected real rate of
return before you make an investment, you can use the promised yield and the expected
inflation rate.
a. Go to www.bankrate.com and click on the CDs and Investments tab. Using
Compare CDs & Investment Rates box, find the average one-year CD rate from
banks across the nation (these will be nominal rates).
b. Use the St. Louis Federal Reserve’s website at research.stlouisfed.org/fred2 as a
source for data about expected inflation. Search for “MICH inflation,” which will
provide you with the University of Michigan Inflation Expectation data series
(MICH). Click on the View Data link and find the latest available data point. What is
the expected inflation rate for the next year?
c. On the basis of your answers to parts (a) and (b), calculate the expected real rate
of return on a one-year CD investment.
d. What does the result tell you about real interest rates? Are they positive or
negative, and what does this mean?
SOLUTIONS TO 5.1 a. The arithmetic average is (2 1 8 2 4)/3 5 2% per month.

CONCEPT b. The time-weighted (geometric) average is
c h e c ks [(1 1 .02) 3 (1 1 .08) 3 (1 2 .04)]1/3 2 1 5 .0188 5 1.88% per month.
c. We compute the dollar-weighted average (IRR) from the cash flow sequence
(in $ millions):
Month
1 2 3
Assets under management at beginning of month 10.0 13.2 19.256

Investment profits during month (HPR 3 Assets) 0.2 1.056 (0.77)
Net inflows during month 3.0 5.0 0.0
Assets under management at end of month 13.2 19.256 18.486
Time
0 1 2 3
Net cash flow* 210 23.0 25.0 118.486
* Time 0 is today. Time 1 is the end of the first month. Time 3 is the end of the third month,
when net cash flow equals the ending value (potential liquidation value) of the portfolio.
The IRR of the sequence of net cash flows is 1.17% per month.
The dollar-weighted average is less than the time-weighted average because the negative
return was realized when the fund had the most money under management.
5.2 a. Computing the HPR for each scenario, we convert the price and dividend
data to rate-of-return data:
www.mhhe.com/bkm
Ending Value Dividend HPR 3 Deviation: Prob 3

Scenario Prob ($ million) ($ million) HPR Prob HPR-mean Dev’n Squared
1 .30 $35 $4.40 .713 .214 .406 .049

2 .45 27 4.00 .348 .157 .040 .001
3 .20 15 4.00 2.174 2.035 2.481 .046
4 .05 8 2.00 2.565 2.028 2.873 .038
Sum: .307 .135
Expected HPR 5 .307 5 30.7%.

Variance 5 .135.
Standard deviation 5 ".135 5 .367 5 36.7%.
5% VaR 5 256.5%.
For the corresponding normal distribution, VaR would be 30.7% 2 1.64485 3 36.7% 5
229.67%.
b. With 36 returns, 5% of the sample would be .05 3 36 5 1.8 observations. The worst
return is 217%, and the second-worst is 25%. Using interpolation, we estimate the
fifth-percentile return as:
217% 1 .8[25% 2 (217%)] 5 27.4%
5.3 a. If the average investor chooses the S&P 500 portfolio, then the implied degree of risk
aversion is given by Equation 5.13:
.10 2 .05
A5 5 1.54
.182
10 2 5
b. S 5 5 .28
18
5.4 The mean excess return for the period 1926–1934 is 3.56% (below the historical
average), and the standard deviation (using n 2 1 degrees of freedom) is 32.55%
(above the historical average). These results reflect the severe downturn of the
great crash and the unusually high volatility of stock returns in this period.
5.5 a. Solving:
1 1 R 5 (1 1 r)(1 1 i) 5 (1.03)(1.08) 5 1.1124
R 5 11.24%
b. Solving:
1 1 R 5 (1.03)(1.10) 5 1.133
R 5 13.3%
5.6 E(r) 5 7 1 .75 3 8% 5
13%
s 5 .75 3 22% 5 16.5%
Risk premium 5 13 2 7 5
6%
Risk premium 13 2 7
Standard deviation 5 16.5 5 .36
www.mhhe.com/bkm

Fin Market Chapter 5

Uploaded by

Copyright:

Available Formats

Fin Market Chapter 5

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fin Market Chapter 5

Uploaded by

Copyright:

Available Formats

Risk and Return:

LO5-1 Compute various measures of return on multi-year investments.

5.1 RATES OF RETURN

Ending price 2 Beginning price 1 Cash dividend

Measuring Investment Returns over Multiple Periods

1st 2nd 3rd 4th

Assets under management at

rG 5 [(1 1 .10) 3 (1 1 .25) 3 (1 2 .20) 3 (1 1 .20)]1/4 2 1 5 .0719, or 7.19% (5.2)

Dollar-weighted RETURN To account for varying amounts under management,

2.1 2.5 .8 .96

Compute the arithmetic, time-weighted, and dollar-weighted average returns.

Conventions for Annualizing Rates of Return

APR 5 [(1 1 EAR)1/n 2 1] 3 n

rcc 5 ln(1 1 EAR) (5.5)

We will return to continuous compounding later in the chapter.

1 1 EAR 5 (1.0101)12 5 1.1282

which implies that EAR 5 .1282 5 12.82%.

5.2 RISK AND RISK PREMIUMS

Scenario Analysis and Probability Distributions

and so the standard deviation is

The Normal Distribution

–4σ –3σ –2σ –1σ 0 +1σ +2σ +3σ +4σ

ri 5 E(ri) 1 sri 3 si (5.9B)

VaR 5 E(r) 1 (21.64485)s (5.10)

3 Normal growth 2 .45 27 4.00

4 No growth 3 .20 15 4.00

5 Recession 4 .05 8 2.00

Normality over Time

Deviation from Normality and Value at Risk

(from Spreadsheet 5.1) to a 3 1.0

Using Time Series of Return

Total 83.4% 886.2

Average rate of return 5 83.4/5 5 16.7

Risk Premiums and Risk Aversion

To determine an investor’s optimal portfolio strategy, we need to quantify his degree of

The Sharpe (Reward-to-Volatility) Ratio

Equity Return Bond Return Long-Term

Total Return—Geometric Average

TABLE 5.2 (concluded)

World Portfolio U.S. Market

Equity Return Bond Return Long-Term

1956–1985 0.34 0.05 0.38 0.28 20.11

5.3 THE HISTORICAL RECORD

U.S. Large Stocks (S&P 500) Long-Term U.S. Treasury

Average Portfolio Return (%)

World Stocks U.S. Small Stocks U.S. Large Stocks

Arithmetic average 10.89 17.57 11.67

Small Stocks T-Bills

Date Small Stocks T-Bills

December 31, 2000 $1,000,000 $1,000,000

5.4 INFLATION AND REAL RATES OF RETURN

The Equilibrium Nominal Rate of Interest

U.S. History of Interest Rates, Inflation,

T-Bills Inflation Real T-Bills

ASSET ALLOCATION ACROSS RISKY

in total returns achieved by institutionally managed pension funds.........There is no reason to

Portfolio Expected Return and Risk

E(rC) 2 rf 5 y[E(rP) 2 rf ] (5.19)

The Capital Allocation Line

Portfolio P: 15% 8% 22% 8