C38FM Topic+3 UtilityTheory
Uploaded by
farid ahmedC38FM Topic+3 UtilityTheory
Uploaded by
farid ahmedSCHOOL OF SOCIAL SCIENCES
Edinburgh Business School
Financial Markets Theory
C38FM
Prof. Mustafa Caglayan
Finance Group
Heriot-Watt University, Edinburgh
Investments and Portfolio Management
Utility Theory
Expected Wealth and Investor Behaviour
In a risky environment investors have to make
decisions, their future wealth is a random variable.
Could we say that investors will always prefer
more expected wealth to less? Do investors care
about any other characteristic of their future
wealth, which is random?
Euler’s game---St Petersburg Paradox
Toss a coin as many times as necessary until it
comes up ‘heads’. How much would you pay to
enter this game?
A head does not appear on the 1st throw 1
A head does not appear until the 2nd throw 2
A head does not appear until the 3rd throw 4
A head does not appear until the 4th throw 8
A head does not appear until the 5th throw 16
A head does not appear until the nth throw 2^(n-1)
St Petersburg Paradox (ctnd)
The probability pay-off matrix of this game is as
follows. What is the expected payoff from this
game?
No of Throws Payoff Probability Payoff x Probability
1 1 2-1 (0.5) 0.5
2 2 2-2 (0.25) 0.5
3 4 2-3 (0.125) 0.5
… … … …
n 2n-1 2-n 0.5
St Petersburg Paradox (ctnd)
Expected return can be calculated as
∞
𝑅ത = 𝐸 𝑅 = 𝑝𝑛 𝑅𝑛
𝑛=0
1 1 1 1 1 𝑛 (𝑛−1)
= 1 + 2 + 4 + 8 + ⋯( ) 2
2 4 8 16 2
1 1 1 1 1
= + + + +⋯ =∞
2 2 2 2 2
• Expected wealth from this gamble is infinitely
large.
Utility: The Expected Utility Hypothesis
One can assign a utility number to each
possible amount of wealth that one might
enjoy.
When these utility numbers are assigned in
a particular way then, given a choice
involving risky alternatives, the individual will
always choose the one that offers the
highest expected utility
The Utility Function (U)
...is a tool to describe the preferences of an
individual, or group of individuals
...does not quantify the amount of utility
received from consumption of the goods
...describes the ranking - it is ordinal and
not cardinal
The Utility Function (U)
We can learn about an individual’s utility
function by observing their choices.
An individual chooses £500 for certain over
a gamble of 50% chance of winning £1,000
and 50% chance of winning nothing, i.e. 0.
U(500) ≷ .5 U(1,000) + .5 U(0)
Observing choices of the individual, we can
then map their utility function. (What is your choice?)
Example
Let’s concern ourselves with only one good -
wealth, denoted w. Our utility function for wealth is
given by
U (w) = 10 + √w
Therefore, if we had a wealth of £100, this would
yield a utility of 20 (utils).
Example2
Simplifying slightly...
Although unrealistic, suppose that we can have no pleasure
without wealth, such that our utility function is given by
U (w) = √w
Lottery
...and we are offered a lottery with the following pay-offs:
a 30% chance of £5,000
a 20% chance of £1,000
a 50% chance of £10
Example 2 (ctnd)
The expected value of the lottery is
.3 (5,000) + .2 (1,000) + .5 (10) = £1,705
The expected utility of this lottery
.3√5,000 + .2√1,000 + .5√10 = 29.12
If this lottery costs £1000 to enter, which do we
prefer?
The lottery yields a utility of 29.12
The utility from £1,000 is √1,000 = 31.62
Certainty Equivalence
So we prefer a certain £1000 to the lottery, even though the
expected value of the lottery is £1,705.
What certain amount of money would make us indifferent between
the lottery and the certain cash?
This is the same question as asking what amount of wealth would
yield the same level of utility as the lottery.
U (w) = √w = 29.12
⇓
w = (29.12)2 = £848
Example 3
Suppose we use a different utility function
U(W) = (W)3/2
Lottery
The utility of the lottery is now given by
.3 (5,000)3/2 + .2 (1,000)3/2 + .5 (10)3/2 = 112,406
Certainty equivalence
The certainty equivalent of the lottery under this utility
function is
112,406 = (w)3/2 => w = (112,406)2/3 = £2,329
Attitude to Risk
Risk averse
Under the first utility function, U (w) = √w, the
individual was indifferent between the lottery, with
an expected value of £1,705, and a certain £848.
Risk-loving
Under the second utility function, U(W) = (W)3/2,
the individual was indifferent between the lottery,
with an expected value of £1,705, and a certain
£2,329.
Risk Aversion
A utility with diminishing marginal utility of wealth.
The investor is risk-averse.
Risk-Loving
• A utility with increasing marginal utility of
wealth. The investor is risk-loving.
Definition: Fair Gamble
• A fair gamble is one where the expected wealth
is same as the choice that is available. Such a
gamble has an expected pay-off of zero. Ex:
A: a £500 gain for certain
B: a 50% chance of winning 1,000 and a 50%
chance of gaining zero.
A risk averse will never accept a fair gamble.
A risk lover will always accept a fair gamble.
Probability Premium
Returning to the utility function U (w) = √w, we offer this
risk-averse individual the choice between a certain £1,000,
or a lottery with equal probability of £2,000 of nothing.
The expected value of the lottery is
.5 (2,000) + .5 (0) = £1,000
What probability, p, of a £2000 win would make the
individual indifferent between the lottery and the certain
payoff?
p√2,000 + (1 − p)√0 = √1,000
Solving for p
P √2,000 + (1 − p)√0 = 31.62 → p = .707
Utility Functions
We know that any positive, affine transformation of the
utility function does not alter the preference ordering.
Subtract from the utility function a constant. (or
multiply with a constant)
Utility is therefore gained from changes in wealth
(return)
Axioms of Rational Choice
1. Comparability: For any pair of assets A and B,
an investor can say A>B; B>A or A=B
2. Transitivity: preferences are transitive; if A>B,
B>C then A>C
3. Independence: If A>B, then a gamble with
outcomes containing A is preferred to the same
gamble containing B
4. For any bet, there is a wealth certain which is
equally preferred.
Utility Functions in terms of Risk and Return
Indifference curves
These preferences can be represented in risk-
return space using indifference curves
Definition
An indifference curve is the locus of all assets (or
combination of assets) between which we are
indifferent.
ҧ σ 𝑝𝑖 𝑥
Recall: E(x) = 𝑥=
Var(x) = σ 𝑝𝑖 (𝑥𝑖 − 𝑥)ҧ 2
Let x be a random variable, then
Var (x) = E(x - E(x))2
= E(x2 – 2xE(x) + E(x)2)
= E(x2) – 2E(x) E(x) + E(x)2
= E(x2) – E(x)2
Shuffling the terms we can write
E(x2) = Var(x) + E(x)2 = Var(x) + 𝑥ҧ 2
Utility Functions in terms of Risk and Return
Risk-averse utility function
We can write a risk averse utility function as a
quadratic utility function in terms of the return, 𝑟𝑥 , on
the asset X.
U (R) = 200𝑟𝑥 − 100 𝑟𝑥2
E (U (R)) = E(200𝑟𝑥 − 100𝑟𝑥2 )
E (U (R)) = 200E (𝑟𝑥 ) − 100E(𝑟𝑥2 )
E (U (R)) = 200E (𝑟𝑥 ) − 100Var(𝑟𝑥 ) – 100E(𝑟𝑥 ) 2
Indifference curves – example
U (R) = 200𝑟𝑥 − 100 𝑟𝑥2
Asset X
E [𝑟𝑥 ] = 0.10
σ [𝑟𝑥 ] = 0.10
Expected utility of asset X
E [U (𝑟𝑥 )] = 200(0.1) − 100 (0.01) − 100 (0.01) = 18
Asset Y
E [𝑟𝑦 ] = 0.12
σ [𝑟𝑦 ] = 0.213
Expected utility of asset Y
E [U (𝑟𝑦 )] = 200(0.12) − 100 (0.0456) − 100 (0.0144) = 18
Indifference curves - example
Which do we prefer?
We are indifferent, i.e. they lie on the same indifference curve.
The expected utility from the two assets, X and Y, is the same.
Asset Z
E [𝑟𝑧 ] = 0.12
σ [𝑟𝑧 ] = 0.16
Expected utility of asset Z
E [U (𝑟𝑧 )] = 200(0.12) − 100 (0.0256) − 100 (0.0144) = 20
Which do we prefer?
We are no longer indifferent. Asset Z yields higher utility than
either of assets X and Y and therefore lies on a higher utility
curve
Indifference curves - example
Increasing risk aversion
