Chapter 1

Comparison of Random Variables.

Preferences of Individuals

probability distribution. Then X1 and X2 have the same expected value m = E{Xi } and
variance σ2 = Var{Xi }.
Ann and David decide to divide the total income into equal shares, so each will have the
random income
1
Y = (X1 + X2 ).
2
Then
1
E{Y } = (m + m) = m,
2
that is, the same as before sharing the risk. On the other hand, the variance

1 2 σ2
Var{Y } = (σ + σ2 ) = ,
4 2
is half as large.
Now, consider n participants of a mutual risk exchange, and denote their random incomes
by X1 , ..., Xn . Assume the X’s to be i.i.d., m = E{Xi } and σ2 = Var{Xi }.
Then the income for each is
X1 + ... + Xn
Y= .
n

σ2
E{Y } = m, while Var{Y } = ,
n
and for large n, the variance is close to zero.
Also,  
X1 + ... + Xn
P → m = 1.
n
We also right the same as
X1 + ... + Xn a.s.
→ m.
n

Thus, for a large number of participants, the risk of each separate participant
may be reduced nearly to zero.

1
2 1. COMPARISON OF RANDOM VARIABLES

1 A GENERAL FRAMEWORK AND FIRST CRITERIA

1.1 Preference order
Suppose you do a job for which you should be paid $50. Your employer suggest to pay
(r.v.) 
 100 with probability 1/2,
ξ=

0 with probability 1/2.
So,
E{ξ} = 50,
the same as the original pay.
Suppose you buy auto insurance against a possible future loss ξ. Assume that with prob-
ability 0.9 the r.v. ξ = 0 (nothing happened), and with probability 0.1, the loss ξ is uniform
on [$0, $2000]. Then
E{ξ} = 0.1 · 1000 = 100.
If the premium c = $110, it means that the loss of ξ is worse for you than the loss of the
certain amount c = 110.
The insurance company: it gets your premium c, and it will pay you a random payment
ξ. The company compares the r.v. Xe = c − ξ with the r.v. Y = 0, and ... .
In the general framework, we consider a class of r.v.

X = {X}
and define a rule of comparison of these r.v.; that is, for each pair (X,Y ) of r.v.’s, we
determine whether X is better than Y , or X is worse than Y , or these two random variables
are equivalent for us.
We may also say that X is not worse than Y , etc.
Formally, we write this as
X % Y,
X ≺ Y,
Y ≻ X,
If simultaneously X % Y and Y % X, we write X ≃ Y .
The monotonicity property:

If X,Y ∈ X and P(X ≥ Y ) = 1, then X % Y . (1.1.1)

The strict monotonicity property:

If X,Y ∈ X , P(X ≥ Y ) = 1, and P(X > Y ) > 0, then X ≻ Y . (1.1.2)

 1. A General Framework and First Criteria 3

EXAMPLE 1. Let two r.v.’s, X = X(ω) and Y = Y (ω), be defined on a sample space Ω
consisting of only two outcomes: ω1 and ω2 , and

X(ω1 ) = 1, X(ω2 ) = 3,
Y (ω1 ) = 1, Y (ω2 ) = 2.

Let V (X) be a function taking on numerical values. We say that an order % is preserved,
or completely characterized, by V (X) if for any X,Y ,

X % Y ⇔ V (X) ≥ V (Y ).

If X,Y ∈ X and P(X ≥ Y ) = 1, then V (X) ≥ V (Y ).

If X,Y ∈ X , P(X ≥ Y ) = 1, and P(X > Y ) > 0, then V (X) > V (Y ).

4 1. COMPARISON OF RANDOM VARIABLES

1.2 Several simple criteria

We will talk about preferences of economic agents—separate individuals, companies,
etc.—using also, for brevity, the term “investor”.

1.2.1 The mean-value criterion

X % Y ⇔ E{X} ≥ E{Y }.
In this case, V (X) = E{X}.
Note that from the mean-value criterion’s point of view, for example, r.v.’s

100 with probability 1/2
X= , and Y = 50 (1.2.1)
0 with probability 1/2

are equivalent.
 1. A General Framework and First Criteria 5

1.2.2 The mean-variance (or mean-standard deviation) criterion

Consider an investor expecting a random income X. Set mX = E{X}, σ2X = Var{X}.
Suppose that the quality of X for the investor is determined just by a function of mX and
σX . In a simplest case,
2

V (X) = τmX − σX , (1.2.2)

where the minus reflects the fact that the quality decreases as the variance increases.
The positive parameter τ is usually called a tolerance to risk.
Note also that often, instead of (1.2.2), people consider V (X) = τmX − σ2X , but the differ-
ence is non-essential.
The function V (X) in (1.2.2) preserves the corresponding preference order % among
r.v.’s:
X % Y ⇔ τmX − σX ≥ τmY − σY .

The mean-variance criteria are very popular, especially in Finance, and at first glance
look quite natural. However, there are situations where the choice of such criteria may
contradict common sense.
EXAMPLE 1. Let X = 0, a number a ≥ 1, and

a with probability 1a ,
Y=
0 with probability 1 − 1a .

Clearly,
E{Y } = 1, Var{Y } = E{Y 2 } − (E{Y })2 = a − 1.

Then,
√ √
V (Y ) = τ · 1 − a − 1 = τ − a − 1, while V (X) = τ · 0 − 0 = 0.

So, whatever τ is, we can choose a sufficiently large a for which V (Y ) < 0. On the
other hand, V (X) = 0, and under the mean-variance criterion, Y is worse than X, whereas
P(X ≤ Y ) = 1. 
EXAMPLE 2. Let X take on values from [1, ∞), and P(X > x) = 1/xα for all x ≥ 1 and
some α > 2. This is a version of the Pareto distribution.
α α
mX = , and σ2X = .
α−1 (α − 2)(α − 1)2

Let Y be uniformly distributed on [0, 1]. Obviously, X ≥ Y with probability one.

√
α α 1 1
V (X) = τ · −√ , and V (Y ) = τ · − √
α−1 α − 2 (α − 1) 2 12
Consequently, for any τ, we can choose α (and hence a r.v. X) such that V (X) < V (Y ).
Thus, under the mean-variance criterion, X is worse than Y , and consequently the crite-
rion (1.2.2) is not monotone. 
6 1. COMPARISON OF RANDOM VARIABLES

One may observe the same phenomenon for V (X) equal to almost
any function g(mX , σX ) of the mean and standard deviation.
Moreover, for any r.v. X, we may point out a r.v. Y such that P(Y ≥ X) = 1
whereas V (Y ) < V (X).

Nevertheless, it is worth emphasizing that the reasoning above does not mean
that we should not use mean-variance criteria, but it does mean that we should
be cautious .
 1. A General Framework and First Criteria 7

F(x) F(x) F(x) F(x) F(x)

1 1 1 1 1
γ γ γ γ γ

0 qγ x 0 qγ x 0 qγ x 0 qγ x x
qγ =0

(a) (b) (c) (d) (e)

FIGURE 1. Quantiles.

1.2.3 Value-at-Risk (VaR)

We adopt the following definition.
Consider a r.v. X with a d.f. F(x) = P(X ≤ x). Let γ ∈ [0, 1].
If the r.v. X is continuous and its distribution function is strictly increasing, then qγ , the
γ-quantile of X, is the unique number q for which F(q) = γ; see Fig.1a.
If there are many numbers q for which F(q) = γ, we take the right end point of the
interval where F(x) = q; see Figures 1b,c.
If the r.v. takes on some values with positive probabilities (and hence the d.f. has “jumps”),
it may happen that there is no number q such that F(q) = γ; see Fig.1d. Then we choose
the point at which F(x) “jumps” over the level γ.
In particular, if X = 0 with probability one, the point 0 is the γ-quantile for all γ ∈ [0, 1)
(see Fig.1e).
Formally, the above definitions may be unified as

qγ = sup{x : F(x) ≤ γ}.

We come to the criterion under discussion. Another term in use is the capital-at-risk
criterion.
Let γ be a fixed level of probability, viewed as sufficiently small. Assume that an investor
does not take into consideration events whose probabilities are less than γ. Then, for such
an investor the worst, smallest conceivable level of the income is qγ .
Let, for instance, γ = 0.05. Then q0.05 is the smallest value of the income among all
values which may occur with 95% probability. One may say that q0.05 is the value at 5%
risk. Note that qγ may be negative, which corresponds to losses.
The VaR criterion is defined as

X % Y ⇔ qγ (X) ≥ qγ (Y ),

i.e., we set V (X) = qγ (X).

In applications of VaR, for the γ-quantile of X, the notation VaRγ (X) is frequently used;
we will keep the notation qγ (X).
8 1. COMPARISON OF RANDOM VARIABLES

EXAMPLE 1. Let a r.v. X (say, a random income) take on values 0, 10 with proba-
bilities 0.1 and 0.9, respectively, and let a r.v. Y take on the same values with respective
probabilities 0.07 and 0.93. The reader is suggested to check that for γ = 0.05, we have
qγ (X) = qγ (Y ) = 0 (look, for example at Fig.1d). So, X and Y are equivalent under the VaR
criterion. However, for γ = 0.08 we have qγ (X) = 0 while qγ (Y ) = 10, that is, X is worse
than Y . So, the result of comparison depends on γ. 
EXAMPLE 2. Let X be normal with mean m and variance σ2 . Since the d.f. of X is
 
x−m
Φ ,
σ

the γ-quantile of X is a solution to the equation

 
q−m
Φ = γ.
σ

Denote by qγs the γ-quantile of the standard normal distribution, i.e.,

Φ(qγs ) = γ.

Then we can rewrite the equation mentioned as

q−m
= qγs ,
σ
and
qγ (X) = m + qγs σ.
The coefficient qγs depends only on γ.
Usually people choose γ < 0.5, and in this case qγs < 0. For example, if γ = 0.05, then
qγs ≈ −1.64 , and the VaR criterion is preserved by the function

qγ (X) ≈ m − 1.64σ.

Criteria of the type

V (X) = m − kσ, (1.2.3)
where k is a positive number. This criterion non-essentially differs from the mean-standard
deviation criterion, but in this case, we do not face any contradiction because we are dealing
only with the normal distribution.
 1. A General Framework and First Criteria 9

EXAMPLE 3. There are n = 10 assets with random returns X1 , ..., Xn . The term “return”
means the income per $1 investment. For example, if the today price of a stock is $11,
11
while the yesterday price was $10, the return for this one-day period is 10 = 1.1. Note that
a return X may be less than one, and in this case we face a loss.
Assume X1 , ..., Xn to be independent and their distributions to be closely approximated
by the normal distribution with mean m and variance σ2 .
Let us compare two strategies of investing n million dollars: either investing the whole
sum in one asset, for example, in the first, or distributing the investment sum equally be-
tween n assets. We proceed from the VaR criterion with γ = 0.05.
For the first strategy, the income will be the r.v. Y1 = nX1 = 10X1 . The mean E{Y1 } = nm,
and Var{Y1 } = n2 σ2 , so to compute qγ (Y1 ) we should replace in (1.2.3) m by nm, and σ by
nσ. Replacing qγs by its approximate value −1.64, we have

qγ (Y1 ) = mn − 1.64nσ = 10m − 16.4σ.

For the second strategy, the income is the r.v. Y2 = X1 + ... + Xn . Hence, E{Y2 } = nm,
Var{Y2 } = nσ2 , and √
qγ (Y2 ) = mn − 1.64 nσ ≈ 10m − 5.2σ.
Thus, the second strategy is preferable, which might be expected from the very beginning.
Nevertheless, when the Xi ’s have a distribution different from normal, we may jump to a
different conclusion.
10 1. COMPARISON OF RANDOM VARIABLES

2 EXPECTED UTILITY
2.1 St. Petersburg’s paradox
The problem below was first investigated by Daniel Bernoulli in a paper published in
1738.
Consider a game of chance consisting of tossing a regular coin until a head appears.
Suppose that if the first head appears right away at the first toss, the payment equals 2 units
of money. If the first head appears at the second toss, the payment equals 4, and so on;
namely, if a head appears at the first time at the kth toss, the payment equals 2k .
It is easy to see that the payment in this case is a r.v. X taking values
2, 4, 8, ..., 2k , ... with probabilities
1 1 1 1
, , , ..., k , ... , respectively, and
2 4 8 2
1 1 1
E{X} = 2 · + 4 · + 8 · + ... = 1 + 1 + 1 + ... = ∞.
2 4 8
By the LLN, this means that if the game is played repeatedly, and X j is the payment in the
jth game, then with probability one
X1 + .. + Xn
→ ∞ as n → ∞.
n
Thus, in the long run, the average payment will be greater than an arbitrary large number.
Then if a player had proceeded from the LLN, she/he would have agreed to pay any,
arbitrary large, entry price for participating in each play.
 2. Expected Utility 11

2.2 Expected utility maximization (EUM)

2.2.1 Utility function
D. Bernoulli proceeded from the simple observation that the “degree of satisfaction” of
having capital, or in other words, the “utility of capital”, depends on the particular amount
of capital in a nonlinear way.
To model this phenomenon, D. Bernoulli assumed that the satisfaction of possessing a
capital x, or the “utility” of x, may be measured by a function u(x) that, as a rule, is not
linear. Such a function is called a utility function, or a utility of money function.
D. Bernoulli himself suggested as a good candidate for the “natural” utility function
u(x) = ln x, assuming that the increment of the utility is proportional not to the absolute
but to the relative growth of the capital. More specifically, if capital x is increased by a
small dx, then the increment of the utility, du(x), is proportional to dx/x, that is,

dx
du = k (2.2.1)
x
for a constant k. The solution to this equation is u(x) = k ln x + C, where C is another
constant. We will see soon that the values of k and C depend just on the choice of units in
measuring utility, and hence do not matter.
Consider now a random income X. In this case, the utility of the income is the r.v. u(X).
Bernoulli’s suggestion was to proceed from the expected utility E{u(X)}.
EXAMPLE 1. Assume that the utility function of the player in St. Petersburg’s paradox
is u(x) = ln x. Then the expected utility
∞ ∞ ∞
E{u(X)} = ∑ u(2k )2−k = ∑ ln(2k )2−k = (ln 2) ∑ k2−k = 2 ln 2,
k=1 k=1 k=1

and, unlike E{X}, the expected utility is finite.

Next, we consider the general case. Clearly, we can restrict ourselves to non-decreasing
utility functions, which reflects the rule “the larger, the better or at least not worse”.
12 1. COMPARISON OF RANDOM VARIABLES

2.2.2 Expected utility maximization criterion

X % Y ⇔ E{u(X)} ≥ E{u(Y )} (2.2.2)

for a utility function u. In particular,

if E{u(X)} = E{u(Y )}, we say that X ≃ Y,

X is equivalent to Y .
The investor who follows (2.2.2) is called an expected utility maximizer (EUM).
The first property of EUM criterion. The preference order (2.2.2) does not change
if u(x) is replaced by any function u∗ (x) = bu(x) + a, where b is a positive and a is an
arbitrary number.
EXAMPLE 1. Consider u(x) = k ln x + C as above. We see now that constants k and C
indeed do not matter, and we may restrict ourselves to u(x) = ln x.
 2. Expected Utility 13
1
EXAMPLE 2. Consider u(x) = − , x ≥ 0 and u∗ (x) =
1 1+x
u (x)
*
x
u(x) + 1 = ; see Fig.2. The second function is positive but
0
x
1+x
u(x) reflects the same preference order as the first. The sign of u(x) does
-1 not matter; what matters when we compare X and Y is whether
E{u(X)} is larger than E{u(Y )} or not. 
FIGURE 2. Now, assume that for a r.v. X, we can find a number c = c(X)
such that c ≃ X with respect to the order %.
The number c(X) so defined is called a certainty equivalent of X.
Now let us consider an EU maximizer with a utility function u. For such a person, the
above relation is equivalent to
E{u(X)} = u(c),

or
c(X) = u−1 (E{u(X)}),

if the inverse exists.

2.2.3 Some “classical” examples of utility functions

1. Positive-power functions: u(x) = xα . for all x ≥ 0 and some α > 0; see Fig.3. The
expected utility in this case is E{u(X)} = E{X α }, the moment of X of the order α.
If α = 1, then E{u(X)} = E{X}, and the EUM criterion coincides with the mean-
value criterion.
For α < 1 the function u(x) is concave (downward), for α > 1 - convex (concave
upward).
For u(x) we are considering, the certainty equivalent of a r.v. X is

c(X) = (E{X α })1/α

. In the simplest case α = 1, the certainty equivalent c(X) = E{X}.

EXAMPLE. Let X be uniform on [0, b]. Then

 1/α
∫b
1 1
c(X) =  xα dx = b.
b (1 + α)1/α
0

Because (1 + α)1/α is decreasing in α, the smaller α, the smaller the certainty equiv-
alent. F
1
Since for α → 0, we have (1 + α)1/α → e, and in this case c(X) = b.
e
14 1. COMPARISON OF RANDOM VARIABLES

u(x) = xα , α < 1 u(x) = −x−α , α > 0

x
0

0 x
(a) (b)
FIGURE 3. Positive- and negative-power utility functions.

2. Negative-power functions: u(x) = −1/xα for all x > 0 and some α > 0; see Fig.3b.
We again deal only with positive r.v.’s, and E{u(X)} = −E{X −α }.
Both cases above may be described by the unified formula
1 1−γ
uγ (x) = x , γ ̸= 1. (2.2.3)
1−γ
In the case γ < 1, we have a positive power function, for γ > 1, we deal with a
negative power function.
3. The logarithmic utility function, u(x) = ln x, x > 0, is in a sense intermediate between
the two cases above and has been already discussed.
4. Quadratic utility functions: u(x) = 2ax − x2 , where parameter a > 0; the multiplier 2
is written for convenience. Certainly, we consider only r.v.’s X such that P(X ≤ a) =
1. Negative values of X are interpreted as the case when the investor loses or owes
money.
We have E{u(X)} = 2aE{X} − E{X 2 } = 2aE{X} + (E{X})2 −Var{X}. Thus, the
expected utility is a quadratic function of the mean and the variance.
5. Exponential utility functions. Let u(x) = −e−βx , where parameter β > 0, and the
function is considered for all x’s.
The graph is depicted in Fig.4.
The expected utility E{u(X)} = −E{e−βX } = −M(−β), where M(z) = E{ezX }.
The function M(z), we also use the notation MX (z) to emphasize that it depends on
the choice of the r.v. X, is the moment generating function of X.
In Exercises, we show that the certainty equivalent
1
c(X) = − ln(MX (−β)).
β

Consider a negative β, setting β = −a for some a > 0. Then

1 1
c(X) = ln(MX (a)) = ln(E{eaX }). (2.2.4)
a a
This is the Masset criterion popular in Economics.
 2. Expected Utility 15

x
0

-1

u(x) = −e−βx

FIGURE 4. The exponential utility function.

In the case of exponential utility, EU maximization has an important property stated in

Proposition 1 Let u(x) = −e−βx and, under the EUM criterion with this utility function,
X % Y . Then w + X % w +Y for any number w.

The number w above may be interpreted as the initial wealth, and X and Y as random
incomes corresponding to two investment strategies. Proposition 1 claims that in the expo-
nential utility case, the preference relation between X and Y does not depend on the initial
wealth.
Proof of Proposition 1. By definition,

w + X % w +Y

iff
E{u(w + X)} ≥ E{u(w +Y )}.
For the particular u above,

E{u(w + X)} = −E{e−β(w+X) } = −e−βw E{e−βX },

and the same is true for Y . So, in the last inequality, the common multiplier −e−βw cancels
out, and the validity of the relation E{u(w + X)} ≥ E{u(w +Y )} does not depend on w.
Hence, if this relation is true for w = 0, it is true for all w. 
16 1. COMPARISON OF RANDOM VARIABLES

2.3 Utility and insurance

Consider an individual with a wealth of w, facing a possible random loss ξ. Assume that
the individual is an EU maximizer with a utility function u(x). What premium G would the
individual be willing to pay to insure the risk?
The individual’s wealth after paying the premium is X = w − G, while if she/he does not
buy the insurance, the wealth will equal the r.v. Y = w − ξ.
In accordance with the principle (2.2.2), a premium G will be acceptable for the person
under consideration only if
u(w − G) ≥ E{u(w − ξ)}. (2.3.1)
For the maximal accepted premium Gmax ,

u(w − Gmax ) = E{u(w − ξ)}. (2.3.2)

(Gmax exists if, say, u is continuous and increasing; we skip formalities.)

EXAMPLE 1. Let u(x) = 2x − x2 , w = 1, and let ξ be uniformly distributed on [0, 1].
Because w − ξ = 1 − ξ ≤ 1, we deal only with x’s for which u(x) increases. Let y = w − G.
By (2.3.1),
2y − y2 ≥ 2E{(1 − ξ)} − E{(1 − ξ)2 }.
Observing that 1 − ξ is also uniformly distributed on [0, 1] (show it!), we have

1 1 2
2y − y2 ≥ 2 − = .
2 3 3
For this inequality to be true, we should have y ≥ 1 − √13 .
Hence, any acceptable premium G ≤ √1 ,
3
and Gmax = √1
3
≈ 0.57. .
 2. Expected Utility 17

Next, we consider not an insured but an insurer. The latter offers the complete coverage
of a loss ξ for a premium H which, in general, may be different from G above. Assume that
the insurer is an EU maximizer with a utility function u1 (x) and a wealth of w1 . (Actually,
it is more natural to interpret w1 as an additional reserve kept by the insurer to fulfill its
obligations.) Following a similar logic, we obtain that an acceptable premium H for the
insurer must satisfy the inequality

u1 (w1 ) ≤ E{u1 (w1 + H − ξ)}, (2.3.3)

and hence for the minimal accepted premium Hmin

u1 (w1 ) = E{u1 (w1 + Hmin − ξ)}. (2.3.4)

EXAMPLE 2. Let u1 (x) = xα , w1 = 1, and ξ be the same as in Example 1. Taking again

into account that η = 1 − ξ is uniformly distributed on [0, 1], we derive from (2.3.3) that
∫ 1
1
1 ≤ E{(H + 1 − ξ)α } = E{(H + η)α } = (H + x)α dx = [(H + 1)α+1 − H α+1 ].
0 α+1
Hence, Hmin is a solution to the equation

(H + 1)α+1 − H α+1 = α + 1.

For example, when α = 1/2, it is easy to calculate—using even a simple calculator— that
Hmin ≈ 0.52. 

For a premium P to be acceptable for both sides, the insurer and the insured, we should
have
Hmin ≤ P ≤ Gmax .
Hence, if Hmin > Gmax , insurance is impossible. If Hmin ≤ Gmax , the premium will be
chosen from the interval [Hmin , Gmax ]. For instance, in the situation of Examples 1-2, we
have 0.52 ≤ P ≤ 0.57.
18 1. COMPARISON OF RANDOM VARIABLES

2.4 Risk aversion

2.4.1 A definition
Consider a r.v. 
ε with probability 1/2,
Zε =
−ε with probability 1/2,
where ε > 0.
Condition Z: For any r.v. X, any ε > 0, and any r.v. Zε independent of X, it is true that
X % X + Zε .
It is important to emphasize that

Condition Z concerns an arbitrary preference order, not only the EUM criterion.

An individual whose preference order satisfies Condition Z is called a risk averter. If

X - X + Zε for any X, any ε > 0, and any Zε independent of X, then we call such an
individual a risk lover or risk taker.
Formally, the above definition does not exclude the case when an individual is simulta-
neously a risk averter and a risk lover, that is, X ≃ X + Zε for all X and ε. In this case, we
say that the individual is risk neutral.
Next, we consider the EUM criterion and figure out when this particular criterion satis-
fies Condition Z.

Proposition 2 Let % be a EUM order defined in (2.2.2). Then Condition Z holds iff
u(x) is concave.

Usually we deal with smooth utility functions, so to check whether an EU maximizer

with a utility function u is a risk averter, it suffices to check the second derivative u′′ .
For example, for u = xα , we have u′′ (x) = α(α − 1)xα−2 . Thus, u′′ (x) < 0 for α < 1,
which corresponds to the risk aversion case, while for α > 1 we deal with a risk lover.
The case α = 1 when E{u(X)} = E{X} may be assigned to both types: the person is risk
neutral.
There is also strong evidence based on experiments that u(x)
many people incline to behave as risk averters when con-
cerned with future gains (positive values of X), and as risk
lovers when facing losses. 0
For example, a person may choose $500 for sure rather x
than $1, 000 with probability 1/2. However, the same person
may prefer to take a risk of losing $1, 000 with probability
1/2 rather than to lose (only) $500 for sure. A utility function
in this case may look as in Fig.5. FIGURE 5.
The following inequality clarifies why the concavity of utility functions is relevant to risk
aversion.
 2. Expected Utility 19

2.4.2 Jensen’s inequality

We assume all expectations below to be finite.

Proposition 3 Let a r.v. X take on values from a finite or infinite interval I, and let a
function u(x) be concave on I. Then

E{u(X)} ≤ u(E{X}). (2.4.1)

If u is convex (concave upward), then

E{u(X)} ≥ u(E{X}). (2.4.2)

Inequalities (2.4.1)-(2.4.2) are relevant to the basic question of insurance: why is it pos-
sible?
If the client is a risk averter (which is natural to assume since the client is willing to pay
to insure a risk), then u(x) is concave, and by Jensen’s inequality,

u(w − Gmax ) = E{u(w − ξ)} ≤ u(E{w − ξ}) = u(w − E{ξ}).

Since u is non-decreasing, it implies that w − Gmax ≤ w − E{ξ}, or

Gmax ≥ E{ξ}.

Thus, the maximum premium the client agrees to pay is larger than (or, for the boundary
case, equals) the average coverage of the risk, E{ξ}.
So, the company will get on the average more than it will pay, which means that the
company can function.
To the contrary, if the client had been a risk lover, from Jensen’s inequality it would have
followed that Gmax ≤ E{ξ}, and insurance would have been impossible.
20 1. COMPARISON OF RANDOM VARIABLES

3 OPTIMAL PAYMENT FROM THE STANDPOINT OF AN IN-

SURED
An individual with a wealth of w is facing a random loss X with a mean m > 0.
The insurer, having many clients, when specifying the corresponding premium g, pro-
ceeds merely from the mean value of the future payment.
As a particular example, we may consider the situation when if the mean payment is λ,
the insurer agrees to sell the coverage for the premium g = (1 + θ)λ for a fixed θ > 0. The
coefficient θ is called a relative security loading coefficient.
For instance, if θ = 0.1, the insurer adds 10% to the mean payment.
If the coverage is complete, the mean payment is equal to the mean loss, that is, λ = m.
However, the individual may wish to buy a non-complete coverage with a mean payment
λ < m.
In this case, the policy is specified by a payment function r(x), the amount that the insurer
will pay if the loss X assumes a value x. Since the coverage is not complete, 0 ≤ r(x) ≤ x.
Note right away, that this implies that 0 ≤ r(0) ≤ 0; that is r(0) = 0.
Thus, the insurer requires only one condition on r(x) to hold:

E{r(X)} = λ. (3.1)

Once the premium g is completely specified by λ, the individual can choose any r(x)
provided that (3.1) is true.
EXAMPLE 1 (Proportional insurance). In this case,

r(x) = kx,

k ≤ 1. Then, E{r(X)} = E{kX} = km, and (3.1) implies that k = λ/m.

Before considering the next example of possible payment functions, let us recall the
following general fact.
 3. Optimal Payment from the Standpoint of an Insured 21
y
1

x
0

Consider a positive r.v. X, so its d.f. F(x) = 0 for x < 0. Let u(x) be a differentiable
function such that the integral ∫ ∞
u(x)dF(x)
0
exists and is finite. Set, also for simplicity,

u(0) = 0.

Then
∫ ∞
E{u(X)} = (1 − F(x))du(x). (3.2)
0
Setting u(x) = x, we obtain from (3.2) a useful formula for the expected value of a posi-
tive r.v.: ∫ ∞
E{X} = (1 − F(x))dx. (3.3)
0
In particular, from (3.3) it follows that the expected value equals the area between the
graph of F(x) and the line y = 1 .
22 1. COMPARISON OF RANDOM VARIABLES

So, (3.1) may be rewritten as follows.

Assume that r(x) is non-decreasing, and set F0 (x) be the d.f. of X. Then
∫ ∞
E{r(X)} = (1 − F0 (x))dr(x).
0

Thus, condition (3.1) may be rewritten as

∫ ∞
(1 − F0 (x))dr(x) = λ. (3.4)
0

EXAMPLE 2 (Excess-of-loss or stop-loss insurance). We will call it also insurance with

a deductible. In this case,

0 if x ≤ d,
r(x) = rd (x) = (3.5)
x − d if x > d,

where the number d is called a deductible. In this case, payment is carried out only if the
loss exceeds the level d, and if it happens, the insurer pays the overshoot.
Inserting (3.5) into (3.4), we have
∫ ∞
(1 − F0 (x))dx = λ. (3.6)
d

The last relation is an equation for d given λ.

Suppose for example that X is exponential with a mean of m. In this case, F0 (x) =
1 − e−x/m , and from (3.6) it follows that
∫ ∞
λ= e−x/m dx = me−d/m ,
d

and m
d = ln m.
λ
Assume, for example that λ = 21 m. Then So, λ = m/2, so d = m ln 2 ≈ 0.69m.
EXAMPLE 3 (Insurance with a limit coverage). In this case,

x if x ≤ s,
r(x) =
s if x > s,

where s is the maximum the insurer will pay. Again using (3.4), we see that restriction (3.1)
may be written as ∫ s
(1 − F0 (x))dx = λ,
0
which is an equation for s. 
We return to the optimization problem. Assume individual to be EUM, and her/his pref-
erences are preserved by ∫ ∞
U(F) = u(x)dF(x), (3.7)
0
and u is a non-decreasing utility function.
 3. Optimal Payment from the Standpoint of an Insured 23

Denote by F(r) (x) the distribution function of the r.v.

X(r) = w − g − X + r(X),

the wealth of the individual under the choice of a payment function r(x). Our goal is to
find a function r for which F(r) is the best. More rigorously, we a looking for a function r∗
which maximizes the function Q(r) = U(F(r) ).
24 1. COMPARISON OF RANDOM VARIABLES

Let us state it rigorously. For a fixed λ ∈ (0, m], consider the set of all function r(x)
satisfying (3.4), that is, the set

R λ = {r(x) : E{r(X)} = λ}.

The theorem below belongs to K.Arrow,

Theorem 4 Let u(x) in (3.7) be concave, and r∗ (x) = rd (x), where d satisfies (3.6).
Then for any function r(x) from R λ ,

Q(r) ≤ Q(r∗ ). (3.8)

So,

The optimal payment is the same for any concave utility function.

EXAMPLE 4. Two people facing the same loss X but having different utility functions, say
√
x and ln x, will prefer the same deductible policy and with the same deductible, provided
that they choose the same mean payment λ.