F79PA Portfolio Theory and Asset Models
Tutorial 1 – Utility Theory
Starter Questions
1. Explain in plain English (and in your own words) what the terms non-
satiation and risk averse mean?
2. For each of the following functions, u(x), is u(x) potentially a utility
function (i.e. does it satisfy non-satiation, and risk aversion), and does
it need to be limited to a particular range of values for x?
A u(x) = −5 exp(−2x)
B u(x) = 1 − exp(4x)
C u(x) = x − exp(−x)
D u(x) = (x − 5)3 + 125
E u(x) = − exp x1
For case C calculate the absolute and relative risk aversion.
3. An economic agent uses the log utility function. At time 1 her wealth
will be either 100 or 200 with equal probability.
Will the certainty equivalent, K, of this outcome be greater than, equal
to or less than the mean outcome of 150?
[Hint: Can you use Jensen’s inequality to help you answer this ques-
tion?]
Calculate the certainty equivalent of this outcome for this agent?
4. An investor, A, with current wealth w has utility function uA (x) must
choose between two investments that will increase his wealth by X1
and X2 respectively.
A second investor, B, also has current wealth w and can also choose
between increases of X1 and X2 . Her utility function is uB (x) = a +
buA (x) where b > 0.
Show that the two investors will both choose the same investment if
they both use the expected utility criterion.
1
More Advanced Questions
5. An investor has current wealth 1, all of which must be invested in one
of two assets. Asset A has a total return that is uniformly distributed
on [1.02, 1.08]. The total return from asset B will be 1.03, 1.04, 1.06
or 1.07, each with probability 14 .
The investor has the quadratic utility function
u(x) = x − αx2 ,
where 0 < α < 0.4629. Determine which asset the investor will prefer.
6. An agent’s utility function is
√
u(x) = 2 x.
(a) What type of utility function is u(x) (e.g. exponential, quadratic,
power, ...)?
(b) Calculate the certainty equivalent of the following uncertain in-
crease in wealth:
£2000 with probability 0.5 and £0 with probability 0.5, assuming:
(i) The agent’s current wealth is £0.
(ii) The agent’s current wealth is £1,000.
(c) Comment on your answers to parts (i) and (ii).
7. An investor has utility function u(w) = − exp(−αw). Her current
wealth is £w = 10, 000 and α = 1/10, 000. An investment results in
an investment gain of X where X = +x with probability 0.75 and
X = −2x with probability 0.25. (So her wealth changes to w + x or
w − 2x respectively.)
Calculate the certainty equivalent when x = 10 and x = 1, 000.
2
8. A risk averse person has an exponential utility function
u(x) = 1 − exp(−αx).
Three random payoffs are available for her to choose from. It is found
that she prefers A to B and prefers B to C. Find the possible range of
the value of α.
A P rob B P rob C P rob
5 1/4 7 1 5 1/3
9 3/4 9 2/3
9. A homeowner is considering buying buildings insurance. His current to-
tal wealth is £200,000. Over one year there is a 1% chance of his house
suffering £100,000 of damage, and a 5% chance of suffering £10,000 of
damage.
(a) A large insurer sells many thousands of identical policies to home-
owners like the one above. What premium should they charge per
policy each year if they expect zero profit?
(b) The homeowner has a log utility function. What is the maximum
price he is willing to pay for this insurance?
(c) Comment on the difference between your answers in parts (a) and
(b).
10. Hyperbolic absolute risk aversion (HARA) utility function is defined as
γ
1−γ ax
u(x) = +b , a, b, x > 0.
γ 1−γ
(a) Show that the absolute risk aversion function is of form 1/(cx+d).
Hence, find c and d in terms of a, b and γ.
(b) By considering γ → 1, show that risk neutral (linear) utility func-
tion is a special case of HARA utility function.
3
11. An investor has a power utility function with parameter γ (γ < 1 and
γ 6= 0).
Suppose that the investor can invest the whole of his initial wealth, w >
0, in one of i = 1, . . . , n investments. The total return on investment i
per £1 invested is Ri > 0. (So the initial wealth of w becomes wRi .)
Prove that the level of the investor’s initial wealth, w, does not affect
the investor’s optimal choice of investment using the expected utility
criterion.