Optimal Portfolio Selection

and Asset Allocation

Prof. Ajaya Panda
Optimal Portfolio Selection and Asset Allocation

• According to Markowitz modern portfolio theory, generally investors

invest a fixed amount for a particular time horizon (i.e., holding
period) and reinvest it after maturity. This happens in the case of
investment in a single security.
• Since portfolio contains a series of securities, this decision can be
viewed as equivalent to selecting an optimal portfolio from a set of
probable portfolios. Hence it is referred to as portfolio selection
problem.
• Investors estimate the expected returns of the portfolio w.r.t. to their
respective risk and then select an optimum portfolio.
Optimal Portfolio Selection and Asset Allocation
The main substance in portfolio theory is to find out the possible & feasible portfolio
opportunity. After identifying the best portfolio on the efficient frontier, the next step is
to identify the level of satisfaction the investors receive from the possible investment
opportunities.

The risk-return trade-off of a portfolio determines investor’s perception towards that

portfolio. Here, the job of the analyst is to judge how the risk of a portfolio affects the
investor’s preferences.

Indifference curve (or) utility functions represents an investor’s preference for a

combination of risk & return. This can be explained graphically in a two-dimensional
space.
Utility Curves
Indifference curve for investors

➢ Each IC represents an equal level of satisfaction.

➢ A higher IC provides a higher level of satisfaction over a lower IC.
➢ Each ICs are locus of portfolios who provide equal level of satisfactions/preference to the investors.
➢ IC map presents ICs with an increasing order of satisfaction.
➢ On ICs all the three portfolios (B, C, D) provide an equal level of satisfaction to the investors even
though they are different in expected return and risk.
➢ Two ICs cannot intersect.
➢ Since a rational investor would expect more return for an additional risk, ICs will always be positively
sloped.
➢ In contrast, for a risk lover investor, ICs will be negatively sloped, or will be skewed towards origin.
➢ The degree of slop associated with ICs indicates the degree of risk aversion ness of the individuals.
That is the slope of ICs indicates whether the investor is aggressive or conservative by nature towards
is investment strategy.
 Characteristics of Utility function:
Out of the various investment opportunities, if the outcomes are certain, it
would be easy for the investor to rank his choices. But in general, the
choices are not obvious. In such cases, the utility function helps in ordering
such random alternatives.
What is a Utility Function:
• It is a function of real numbers and gives real values. Once the utility
function “U” is defined, the random wealth levels are ranked on the
basis of their Expected utility values.
Example: E[U(x)] and E[U(y)]
➢ The higher the value, the higher the preference.
➢ Utility functions vary among investors depending on their level of risk
tolerance.
➢ One general restriction on the utility function is that it is an increasing
continuous function. That is if x > y, the U(x) > U(y).
➢ Other than this restriction, the utility function in theory can take any
form.
 Types of Utility Function

(a) Linear utility function: U(x) = X

• This utility function is said to be risk neutral as no account for
risk is made.
(b) Exponential utility function: U(x) = -e-ax for a > 0.

• This utility function has negative value. This negativity does not
matter much, as relative values are important for comparison.
• The exponential utility function is most appropriate for people
whose risk attitude does not change according to the amount of
wealth they have. Many individuals might be less risk-averse if
they had more wealth.
• This utility function will be increasing towards zero.
 Types of Utility Function
(c) Logarithmic Utility Function: U(x) = ln(x) for x > 0, And if x = 0, U(x) = ∞

• The logarithmic utility-of-consequences function has constant relative risk-

aversion; therefore, the fraction of your income allocated to the risky asset is
independent of your initial wealth.

(d) Quadratic Utility Function: U(x) = x – bx2

• The absolute risk aversion function demonstrates that the quadratic utility
function exhibits increasing absolute risk aversion. Thus, the quadratic function
is consistent with investors who reduce the nominal amount invested in risky
assets as their wealth increases. By definition, a quadratic utility function must
exhibit increasing relative risk aversion.

(e) Power Utility Function: U(x) = bxb , for b ≤ 1 and b ≠ 0

• If b = 1, the U(x) will become linear and will represent the utility preference of
risk neutral investor.
Example:

Suppose there are two investment opportunities

for a venture capitalist. [Link] Probability E(R)
➢ Option 1: Buy a T-bill Govt, bond and 1 0.20 $ 10 million
expect a return of $ 6 million with certainty.
2 0.40 $ 05 million
➢ Option 2: Invest in a risky portfolio with a 3 0.40 $ 01 million
certain probability of return distribution.
If the analyst uses a power utility function i.e.,
U(x) = x1/2, then which investment opportunity
he should prefer for the investor.
Utility from option 1:

• U(x) = (6 𝑚𝑖𝑙𝑙𝑖𝑜𝑛) = 2.45

Utility from option 2:

• U(x) = 0.20 × 10 𝑚𝑖𝑙𝑙𝑖𝑜𝑛 + 0.4 × 05 𝑚𝑖𝑙𝑙𝑖𝑜𝑛 + 0.4 × 1 𝑚𝑖𝑙𝑙𝑖𝑜𝑛

= 1.93
Since, the investor gets highest utility from the investment option 1, the analyst should
suggest Option 1 as best choice for him.
Axiomatic Specification of Utility function

➢ Adding a constant to the utility function does not affect its ranking.
Suppose we are using a utility function U(x) at a particular time for a group of investors and an alternate
utility function V(x) = U(x) + b, is used by another analyst. The new utility function V(x) is expected to
generate same ranking as previous U(x)
• 𝐸𝑉 𝑥 = 𝐸 𝑈 𝑥 +𝑏 ⇒ 𝐸𝑈 𝑥 +𝑏
➢ Multiplying a constant to the utility function does not affect its ranking.
• 𝑉 𝑥 =𝑎𝑈 𝑥 for a > 0
• 𝐸𝑉 𝑥 = 𝐸 𝑎𝑈 𝑥 = 𝑎𝐸 𝑈 𝑥

➢ Both adding and multiplying a constant to the utility function does not affect its ranking.
• 𝑉 𝑥 = 𝑎 𝑈 𝑥 + 𝑏 for a > 0 is also a utility function equivalent to U(x) generating identical ranking.
 Risk Aversion and Utility Values
• We begin by introducing two themes of portfolio theory that
are centered on risk.
(1) Tenet: It says investors will avoid risk unless they can
anticipate a reward for engaging in risk.
(2) Personal Utility function: It allows us to quantify investors
personal trade-offs between portfolio risk and expected
return. The utility function allows each investor to assign a
welfare or utility score, based on their expected return and
risk preferences and the choose the portfolio with the highest
score.
• Based on the utility function, fund managers can decide how
much of their wealth to put at risk for the greater expected
return to be achieved.
 Risk Aversion and Utility function
• In the previous discussion, we have seen Portfolio E(R) Risk Risk
that risky assets commands risk premium in Premium (σ)
the market. The risk-averse investor E(R) - Rf
penalizes the expected return of a risky
portfolio by a certain percentage on account 1 L (Low Risk) 07 % 02 % 05 %
of its risk and demands a positive risk 2 M (Medium Risk) 09 % 09 % 10 %
premium. Greater the risk, greater is the
penalty or risk premium. 3 H (High Risk) 13 % 13 % 20 %
Example:
Suppose Risk free rate (Rf) = 5%
 Utility Function
• One could easily rank these portfolio w.r.t risk premium. But
note that the risk of the portfolio is increasing a long with
return may put the investor in an indifferent state.
• Then how does the investor quantify the trade-off?
• There are various scoring models that explain these trade-offs.
But the financial theorist and CFA institute popularly use a
“utility function.”
• Utility Function
1 2
𝑈 = 𝐸 𝑟 − 𝐴𝜎
2
• U = Utility or wealth score of the investor
• A = Index of investor’s risk Aversion
• ½ = Is a factor of scaling conversion.
 Utility Function
➢ The utility function is based on expected return
(E(r)) and risk measured as σ2.
➢ U increases with increase in E(r) and decrease in
σ2 and vice versa.
➢ In case of a risk-free asset, U is equal to E(r) of
the Rf.
1
➢ 𝐴𝜎 2 constitutes the penalty component for the
2
risk.
• A > 0 or (+ve) for Risk Aversion investor
• A = 0 or for Risk neutral investor
• A < 0 or (-ve) for Risk lover investor
 • Let’s calculate the utility score of some alternate
portfolios for investors with different risk aversion
coefficient.
Utility of Portfolio

Utility of Portfolio (L) Utility of Portfolio (M) Utility of Portfolio (H)

A E(r) = 0.07, σ = 0.05 E(r) = 0.09, σ = 0.1 E(r) = 0.13, σ = 0.20

1 2.0 0.07–(½×2×0.052) = 0.0675 0.09 – (½×2×0.12) = 0.08 0.13–(½×2×0.202) = 0.09

2 3.5 0.07–(½×3.5×0.052) = 0.0656 0.09–(½×3.5×0.12) = 0.0725 0.13– (½×3.5×0.22) = 0.06

3 5.0 0.07–(½×5×0.052) = 0.0638 0.09–(½×5×0.12) = 0.065 0.13– (½×5×0.22) = 0.03

 Utility of Portfolio
• Risk aversion is the tendency to avoid risk and have
a low risk tolerance. Risk-averse investors prioritize
the safety of principal over the possibility of a
higher return on their money.
• Here we notice that the high-risk portfolio (H) will
be chosen by the investor with lowest degree of risk
version (i.e., A = 2) and the lower risk portfolio is
passed over by the most risk aversion investors.

• The utility scores are otherwise known as

“Certainty Equivalent Rate” of return. Certainty
Equivalent Rate is the rate that any risk-free
investment would need to offer to provide the same
utility as the risky portfolio. Hence, Certainty
Equivalent Rate is always greater than risk-free rate
with a positive risk premium.
Indifference
Curve