Efficient Frontier (EF), Optimum Portfolio, and Capital Market Theory

1. Efficient Frontier (EF)

If we consider the infinite number of portfolios that could be formed from two or more securities
and plotted these portfolios’ expected return and risk, we would create a graph like the one in
Fig. 1. The efficient frontier is represented by the heavy dark line from E to F in Fig. 1. Portfolios
along curve EF dominate all other investment possibilities. Portfolio F, the highest return
portfolio, is the only one that is likely a one-asset portfolio. The curvature of the efficient frontier
depends upon the correlation of the asset’s returns. The efficient frontier curve is convex toward
the E(r) axis because all assets have correlation coefficients that are less than positive unity and
greater than negative unity.

Figure 1. Efficient Frontier.

2. Optimum Portfolio

Markowitz portfolio analysis makes the following assumptions:

1. The investor seeks to maximize the expected utility of total wealth.

2. All investors have the same expected single period investment horizon.
3. Investors are risk-averse.
4. Investors base their investment decisions on the expected return and standard deviation of
returns from a possible investment.
5. Perfect markets are assumed (e.g., no taxes and no transaction costs).

Given the above assumptions, an investor will want to hold a portfolio somewhere along the
efficient frontier (that is, along EF in Fig. 1). The exact location depends on the investor’s risk-
return preferences. A set of indifference curves for each investor will show his or her risk-return
Trade-off. Those investors with more risk-aversion require more compensation for assuming risk
and will choose a portfolio along the lower end of the efficient frontier close to E in Fig. 1. The
portfolio chosen will be optimal because no other portfolio along the efficient frontier can
dominate another in terms of risk and return.

3. Capital Market Theory

A. Capital Market Line (CML)

The assumptions underlying capital market theory are as follows:

1. Investors are Markowitz efficient diversifiers who delineate and seek to

attain the efficient frontier. (The assumptions listed for the Optimum Portfolio apply.)
2. Money can be borrowed and lent at the risk-free interest rate, denoted R.
Stated differently, var(R) = 0.
3. Investors have homogeneous expectations (idealized uncertainty).
4. Investments are infinitely divisible.
5. No taxes or transaction costs exist.
6. No inflation exists.
7. Capital markets are in equilibrium.

When we introduce a risk-free asset into Markowitz portfolio analysis, given the above
assumptions, the efficient frontier is changed from a curve to a straight line. This new efficient
frontier is called a capital market line (CML) and is illustrated in Fig. 2. The CML starts with the
risk-free asset R and is tangent to the risky portfolio m on the Markowitz efficient frontier. Note
that portfolio m is now the only risky portfolio desired.
 Figure 2. The capital market line and the efficient frontier.

To the left of m, investors on the CML will hold both the risk-free asset and the risky portfolio.
Since these investors are holding part of their investment in R, they are lending at the rate of R.
Just the opposite is true to the right of m. In this case investors are borrowing at R and investing
more in m—they are utilizing leverage. Portfolio m is called the market portfolio and contains all
assets.

The expected return on a CML portfolio composed of R and m is

E(rp) = xRR + xmE(rm)

Or since
xR is the proportion of an investor’s wealth invested in R, and E(rm) is the expected return on the
market portfolio.

The risk of a portfolio composed of R and m is measured with the following quadratic equation:

However, since , then , and the risk formula collapses to the following linear form:

where σpis the standard deviation of the portfolio’s returns and σm is the standard deviation of
the market portfolio’s returns.

Exercise 2.1 Suppose the standard deviation of the market portfolio is 20 percent, its expected
return is 14 percent, and the risk-free asset is 9 percent. What return can an investor expect to
earn on an investment of 50 percent of his wealth in the risk-free asset and 50 percent of his
wealth in the market portfolio? What is the 50-50 portfolio’s risk?

Exercise 2.2 Suppose you are interested in investing 60 percent of your wealth in the
market portfolio and 40 percent in the risk-free asset. What is the expected risk and return of the
60-40 lending portfolio? Assume the market portfolio has an expected return of 15 percent and
a standard deviation of 25 percent.The risk-free rate is 10%.

Exercise 2.3. Refer to Exercise 2.2 and assume the same inputs. If you borrowed at risk so that
xR is −.5 and invested 1.5 times your wealth in the market portfolio, so that , what will be your
expected risk and return? .

Since borrowing is negative investing, the weight for the risk-free rate is negative and xm
exceeds 1.0 so that the weights sum to 1.0. Thus, we have
 B. Simple Diversification Reduces Risk

An asset’s total risk can be divided into systematic plus unsystematic risk, as shown below:

Systematic risk (undiversifiable risk) + unsystematic risk (diversifiable risk) = total risk = var(r)

Unsystematic risk is that portion of the risk that is unique to the firm (for example, risk due to
strikes and management errors). Unsystematic risk can be reduced to zero by simple
diversification. Simple diversification is the random selection of securities that are to be added to
a portfolio. As the number of randomly selected securities added to a portfolio is increased, the
level of unsystematic risk approaches zero. However, market-related systematic risk cannot be
reduced by simple diversification. This risk is common to all securities. Figure 3 illustrates how
total risk approaches systematic risk as the number of securities in a portfolio increases.

Figure 3. Portfolio size, total risk, and systematic risk.

C. Characteristic Line

Systematic risk can be measured statistically by using ordinary least squares (OLS)
simple linear regression analysis. A financial model called the characteristic line is used to
measure both systematic and unsystematic risk.

The equation for the characteristic line (or regression line) is

where ai = the intercept for the ith asset

bi = the slope b for the ith asset, a measure of undiversifiable risk
eit = the random error around the regression line for security i during time
period t
 The above equation shows the relationship of one security with the market and is
sometimes called a market model for one security. OLS regressions are formulated so that the
error terms (eit) will average out to zero. As a result, the characteristic line is normally written
(without the time subscripts) as

The term ai is called an alpha coefficient for security i. It measures the ith asset’s rate of
return when the market return . The term bi is called the beta coefficient; it measures the slope
of the characteristic line. Beta is defined mathematically as

where cov(ri, rm) = the covariance of returns of the ith asset with the market
var (rm) = the variance of the returns of the market index

Exercise 3.1 Using the characteristic line model’s beta, define an aggressive asset. What
is a defensive asset?

The beta coefficient is an index of systematic risk. Betas can be used for an ordinal ranking of
the systematic risk of assets, but not for a direct comparison with total risk or systematic risk. An
asset with is an aggressive asset because it is more volatile than the market portfolio. For
example, if an asset has a beta of 2 and the market (e.g., as represented by the S&P500) goes
up by 10 percent, this asset will increase in return by 20 percent on average. With a defensive
asset, beta is < 1, and the response of the asset will be less than that of the market.

Exercise 3.2 Show mathematically how total risk for an asset can be partitioned into the
systematic and unsystematic risk components.

The unsystematic risk measure, var(e), is called the residual variance (or
standard error squared) in regression terms.

Exercise 3.3 The annual rates of return for the X Corporation and the market are
given below:
YEAR X returns Market returns
1 -5% -6%
2 14 16
3 10 12
4 12 14
5 17 20

(a) Determine the beta coefficient for X.

(b) What percent of the total risk for X is systematic?

Solution:
(a) The same beta coefficient can be calculated with the following computationally efficient
equation:

where N is the number of observations, Xi is the return for security Xi, and Yi is the return for
security Yi

Year rx rm r2m rxrm

1 -5 -6 36 30
2 14 16 256 224
3 10 12 144 120
4 12 14 196 168
5 17 20 400 340
Σ = 48 Σ = 56 Σ = 1,032 Σ = 882

Note: The independent (or explanatory) variable is usually called X, so = the market return. The
dependent variable is = the individual stock return.

With N = 5,

(b) The correlation coefficient squared (ρ²) will tell you what percent of total risk is systematic.
The following equation can be used to calculate ρ:
The above equation was discussed in the previous material. Note that the equation used in (a),
includes most of the values needed to calculate ρ. The exception is the Y² term.

Y2
25
196
100
144
289
754

With Y² and the other inputs from part (a), ρ can be calculated as follows:

D. Capital Asset Pricing Model (CAPM)

CAPM has been discussed already in the previous material. For more illustration, see
the excel file illustrating the computation and application of the CAPM or SML.