BEHAVIORAL FINANCE

LECTURE 1
CHOICE UNDER CERTAINTY

Roman Goncharenko, PhD

KU Leuven

September 28,
2023
Lecture Outline

• Mathematical preliminaries: Relation xRy

• Preference relation
• Rational preferences
• Choice under certainty
• Utility function
• Opportunity costs
• Sunk costs
• Menu dependence and the decoy theory
• Loss aversion and the endowment effect
• Anchoring and adjustment
Mathematical preliminaries: Sets

• Formally, a set is a collection of non-identical elements

• for example consider set A = {a1 , a2 , a3 }, since a2 is an element of A
we write a2 ∈ A
• Consider set A0 = {a1 , a2 }; since every element of A0 is also found in A,
we say that A0 is a subset of A and write A0 ⊂ A
• Consider set B = {b1 , b2 , a3 }; we define the intersection of sets A and
B as
A ∩ B = {x : x ∈ A and x ∈ B}
• from the definition, it follows that A ∩ B = {a3 } since a3 is the only
common element of A and B
• We define the union of sets A and B as

A ∪ B = {x : x ∈ A or x ∈ B}
• from the definition, it follows that A ∪ B = {a1 , a2 , a3 , b1 , b2 }
Mathematical preliminaries: Sets
Mathematical preliminaries: Sets and Cartesian
product
• Let A and B be two non-empty sets, the Cartesian product of A and
B is defined as
A × B = {(a, b) : a ∈ A and b ∈ B}.
• Let A = {a1 , a2 } and B = {b1 , b2 }, then from the definition

A × B = {(a1 , b1 ), (a2 , b1 ), (a1 , b2 ), (a2 , b2 )}

• (convince yourself that, in general, A × B 6= B × A)
• the Cartesian coordinate system is an example of the Cartesian product
Mathematical preliminaries: Relation xRy

• A (binary) relation on a nonempty set X is a subset R ⊆ X × X

• we write
xRy ⇐⇒ (x , y ) ∈ R
• Example:

If X = {x1 , x2 , x3 } then X × X has 32 = 9 elements. Let’s define R such

that R = {(x1 , x1 ), (x2 , x2 ), (x3 , x3 )}. Clearly, R ⊆ X × X and thus
defines a relation on X .
alternatively and equivalently, we can write x1 Rx1 , x2 Rx2 , and x3 Rx3
Note, however, that given R in our example, there is no relation between
any two different elements on X . So, any idea what this relation actually
is (you all know it but might not recognize at this moment)?

• On the next slide, we will list various important properties of relation

Mathematical preliminaries: Relation xRy

Let R be a relation on X and denote by x , y , z generic elements of X . We

say that R is

• reflexive if xRx for all x ∈ X

• symmetric if xRy implies yRx
• antisymmetric if xRy and yRx implies x = y
• transitive if xRy and yRz implies xRz
• complete if for any x , y ∈ X either xRy or yRx (or both)

Example: Let X = {1, 2, 3} and R = {(1, 2), (1, 3), (2, 3)}. Which properties
does this relation satisfy? Do you recognize this relation?
Preference relation
Let X be a set of alternatives. For example, think of alternative ways of
traveling to Brussels from Leuven, in which case X = {bike, bus, car, train}.

Our weak preference relation on the set of alternative X is denoted by

symbol % (before we used a generic R). For any two elements (alternatives)
x , y ∈ X if we write
x % y,
we mean that "x is at least as good as y ."

There are two more preference relations that we can derive from %:

• The strict preference relation . If x  y , we have x % y , but not

y % x . We read x  y as "x is preferred to y."
• The indifference relation v. If x v y , we have x % y and y % x . We
read x v y as "x is indifferent to y."
Rational preference relation

The preference relation % on the set of alternatives X is rational if it is:

(1) complete: for all x , y ∈ X either x % y or y % x (or both)

(2) transitive: for all x , y , z ∈ X if x % y and y % z then x % z

This may not seem like asking too much. However, haven’t you ever found
yourself in a situation when you simply cannot decide between two
alternatives? If you have, then this is example of preference incompleteness.
Likewise, young children sometimes may exhibit intransitive preferences.
Rational preference relation: practice exercise

To get yourself better familiar with preference relation, prove the following
statements:

If the preference relation % is rational, then

(1)  is both irreflexive ( x  x never holds) and transitive
(2) v is reflexive, transitive, and symmetric
(3) if x  y % z, then x  z
Utility function
Ideally, we would like to describe preference relations % on a set of
alternatives X by means of a utility function. The reason is simple:
mathematically, it is much easier (and efficient) to work with functions than
with binary relations. Thus, our utility function, u(x ), will assigns a
numerical value to each element x ∈ X , thereby ranking the elements of X in
accordance with the individual’s preferences.

A function u : X → R is a utility function representing preference relation

% if for all x , y ∈ X
x % y ⇐⇒ u(x ) ≥ u(y )

Note: If such a function exists, then we can work with utility function
instead of preferences. The existence of such a function, in general, is rather
hard to prove (and is far beyond the scope of this course) and is required a
couple of more technical assumptions on our preferences %. What you can
try to prove, however, is that if for any % there exists u : X → R such that
for all x , y ∈ X x % y ⇐⇒ u(x ) ≥ u(y ), then % must be rational.
Additional assumptions on preferences

The consumer-choice theory under certainty is for practical purposes often

supplemented by various additional assumptions.
(1) non-satiation: given any element in x ∈ X close to it there is y ∈ X
such that u(y ) ≥ u(x )
(2) convexity: for any x , y , z ∈ X and for any α ∈ [0, 1] if x % y and
z % y then αx + (1 − α)z % y
(3) continuity: there are no "jumps" in preferences—no "lexicographic"
preferences
Indifference sets/curves

Let x ∈ X . We define indifference sets as a collection of y ∈ X such that

y v x.

Suppose that X = {(x1 , x2 ) : x1 , x2 ∈ R+ }. Then indifference sets are curves

and can look like this:
Budget sets and optimal choice
In general, not all alternatives are obtainable since there are budget
constraints. Budget set B is a subset of X , which includes only those
alternatives which are obtainable.

Graphically, a budget constraint together with the indifference curves may

look like this:

Note, that since we assume that the utility of indifference curves increases as
we move in the North-East direction, the black dot (x˜1 , x˜2 ) represents the
optimal choice–a bundle (element) with the highest utility yet within the
budget set (on the budget line).
Optimal choice: example
Suppose the consumer’s preferences are given by the following utility
function:
u(x ) = ln(x1 ) + ln(x2 )
and the budge set is given by

p1 x1 + p2 x2 ≤ w ,

where p1 and p2 are the prices of goods 1 and 2, respectively, and w is the
available income.

Since preferences are strictly increasing, it makes no sense to leave any

income unspent and, thus, p1 x1 + p2 x2 ≤ w . Then the optimal choice
involves maximizing the utility function subject to the budget constraint,
which produces the following optimal choice over (x1 , x2 ):
!
w w
(x˜1 , x˜2 ) = ,
2p1 2p2
Optimal choice: practice

Next, redo the exercise from the previous slide assuming the preferences are
given by the following utility functions:

u(x ) = x1α x2β ,

u(x ) = ax1 + bx2 ,

u(x ) = max{x1 , x2 }.
Indifference sets: practice

For each of the sets of indifference curves below, what assumptions are
violated?
Discussion
• Is rationality within the context of choice under certainty restrictive?
• All we need for this theory to work is two assumptions for rationality,
which do not seem to be so restrictive
• What this theory does not say is that people are selfish, materialistic, or
greedy; neither does it say anything about why people have the
preferences they do nor that they solve the mathematical optimization
model in their heads.
• Much of the criticism of economics for its rationality assumption is
misguided (not all of it, though—otherwise we would not be here now)
• Is this a plausible theory of human behaviour? A theory is useful if it
helps to explain reality. Whether a theory helps to explain reality can be
tested empirically. Thus, a theory that receives substantial empirical
support can indeed be useful, although, it should never be overlooked
that a theory is still a theory–simplification and it can never correspond
one-to-one to reality.
• Next, we will examine some common behavior patterns that do violate
the theory of rational choice under certainty and learn about basic
alternatives
Opportunity cost (1/4)

Imagine the following rather realistic scenario:

• you invest in real estate and later realize you made some profit
• you are happy and boast about your money-making skills
• someone points out to you that you would have made more (in
expectation) if you had invested in stock
• still you feel like it ws a great idea to invest in real estate because at
least you didn’t lose any money

but this is irrational behavior because it looks like you missed to consider the
opportunity cost of your investment
Opportunity cost (2/4)

Opportunity costs of choosing alternative A are the costs that are due to
the loss of other (mutually exclusive) alternatives that you forgo when
choosing A

What are the opportunity costs of

• living in your own house?
• investing in real estate?
• going for holidays to Portugal?
• being present in this class?
• studying for a university degree?

Failing to account for opportunity costs distorts the preference ordering and
leads to irrational behaviour
Opportunity cost (3/4)

Opportunity costs of choosing alternative A are the costs that are due to
the loss of other (mutually exclusive) alternatives that you forgo when
choosing A

Example:
• you have two choices how to spend the next two weeks:
a) working on a project that yields $800
b) having holidays in Portugal for $2500
• what is the cost of going to Portugal? $2500? Don’t forget the
opportunity cost of $800. Thus, the total cost is actually $3300
• Should you go? This depends on what’s your reservation price for
going to Portugal is
• If you value the trip to Portugal for more (less) than $3300 then you
should (not) go
Opportunity cost (4/4)

The following question was famously asked of 200 professional economists at

the 2005 meeting of the American Economic Association.

You won a free ticket to see an Eric Clapton concert (which has no resale
value). Bob Dylan is performing on the same night and is your next-best
alternative activity. Tickets to see Dylan cost $40. On any given day, you
would be willing to pay up to $50 to see Dylan. Assume there are no other
costs of seeing either performer. Based on this information, what is the
opportunity cost of seeing Eric Clapton? (a) $0, (b) $10, (c) $40, or (d)
$50.

Only 21.6% of the professional economists in the study got the answer right,
which is particularly embarrassing because even if they had answered at
random the result would have been better...
Sunk cost (1/5)

An old and classic example. Consider the following situation:

• you went to a shop and bought 1kg of some very expensive apples
• upon trying them at home, you discovered that they were rather terribly
sour—which, let’s assume, you hate
• nevertheless, you decide that you must proceed eating them otherwise
you’d be losing the money you paid for them
If that (or a similar) scenario has ever happened to you, then you might have
acted irrationally by basing your decision on sunk cost
Sunk cost (2/5)
Sunk cost is a cost that has been already incurred (and cannot be
recovered, thus, "sunk") before the decision at hands has been made

An extended example:
• you are planning a holiday in Portugal and your utility from doing it is
equivalent to $2000
• you book the flight and an AirB&B for $1800 (cancellation policy–no
refund)
• unexpectedly you learn that your airline company goes bankrupt. While
you are fully reimbursed ($700) you still have to buy new tickets if you
are to go
• what is the highest price you will be rationally willing to pay for the new
tickets?
• what is the highest price you will be rationally willing to pay for the new
tickets if the cancellation policy was "50% refund if canceled a day
before the starting date"?
Sunk cost (3/5)
Let X denote the price of new tickets (note: I omit the expense on the old
tickets as it is anyways fully refunded). A decision tree for this problem
under no-refund cancellation policy is:

You should go to Portugal only if a payoff from doing it is higher than that
from not doing it—that is, if

2000 − X − 1100 ≥ −1100

2000 − X ≥ 0
X ≤ 2000
Sunk cost (4/5)
Again, let X denote the price of new tickets. A decision tree for this problem
under 50%–refund cancellation policy is:

Again, you should go to Portugal only if a payoff from doing it is higher than
that from not doing it—that is, if

2000 − X − 1100 ≥ 550 − 1100

2000 − X ≥ 550
X ≤ 1450
and with a full-refund cancellation policy X ≤ 900
Sunk cost (5/5)

• Failing to ignore sunk costs is referred to as honoring sunk cost or

committing the sunk-cost fallacy
• Sometimes it is also called as the Concorde fallacy, since French and
British government continued to fund the supersonic passenger jet long
after it became cleat that it would not be commercially viable
Menu dependence and the decoy effect
The following picture sums it up really well:

Suppose X = {x , y , z}. Let’s say x % y . If you are offered a menu {x , y },

then you rationally choose x . Now, if you offered the menu X , then rationally
you can’t choose y since either z % x , in which case you must choose z, or
x % z, in which case you must choose x . For more, read Section 3.4 in BE
Loss aversion and the endowment effect (1/3)
• In our rational theory so far, the preferences are assumed to be
independent of your endowment
• Suppose your preferences over a bottle of beer are represented by a
utility function u(x ) = ln(x )

• Note that while the bottles of beer always lead to higher utility, the
incremental (marginal) increase with each bottle actually decline
• Marginal utility:
d 1
u(x ) = u 0 (x ) =
dx x
Loss aversion and the endowment effect (2/3)

• In practice, however, people rarely behave in the way the theory predicts:
• frequently, people require a lot more to give up a unit of consumption
they already have, as a part of their endowment, than they would be
willing to pay to purchase a new one
• Such an endowment effect is typically explained as a result of loss
aversion
• Loss aversion is the behavioral trait that people dislike losses more
than they like commensurate gins
• Loss aversion is captured by means of value function ν(.), which
represents how an agent evaluates a change and which is essential part
of Prospect Theory
Loss aversion and the endowment effect (3/3)

• Prospect theory is a theory of behavioral economics and behavioral

finance that was developed by Daniel Kahneman and Amos Tversky in
1979. (The theory was cited in the decision to award Kahneman the
2002 Nobel Memorial Prize in Economics.)
• Based on results from controlled studies, it describes how individuals
assess their loss and gain perspectives in an asymmetric manner (i.e.,
loss aversion)
• We will talk more about prospect theory and loss aversion once we move
to optimal choice under uncertainty
• Read Section 3.5 in BE
Anchoring and adjustment (1/3)

Consider the following show. First, you draw a random number between 0
and 100 and then invite participants yo answer the following two questions:

(a) Is the percentage of African nations in the UN greater or less than the
number?
(b) What is the actual percentage of African nations in the UN?

You would not expect the answer to (b) to reflect the random generated
number. Yet when this experiment was performed, a correlation between the
two was found. When the starting point was 10, the median answer to (b)
ws 25; when the starting point was 65, the median answer was 45.

This phenomenon is often explained by reference to anchoring and

adjustment
Anchoring and adjustment (2/3)
• Anchoring and adjustment is a cognitive two-stage process that can
be used when forming judgments
• first, you pick an initial estimate called anchor
• second, you adjust the initial estimate up or down (as you see it fit) in
order to come up with a final answer
• According to the heuristics-and-biases program, we make judgments
not by actually computing probabilities and utilities but by following
heuristics
• A heuristics is a rule of thumb or mental short cut that can be used
when forming judgments
• The functionality of heuristics is that they reduce the time and effort
required to solve everyday problems and produce an approximate solution
• heuristics may lead to a bias, the situation in which heuristics produce a
systematical and predictable wrong answer
• Read more in Section 3.6 in BE
Anchoring and adjustment (3/3)

• Anchoring and adjustment can be seen in many situations in finance

• For example, one may get anchored to the result of a valuation model
and make decisions or negotiate around it
• It ignores the model error that arises from incorrect assumptions or if the
model is suitable to begin with
• Sometimes, people may be anchored to figures in a plan or a forecast
that may not be relevant to the current situation
Discussion
• In this lecture, we introduced the theory of a rational choice under
certainty
• We discussed that while rationality does not require a lot in terms of
assumptions, in practice people often behave irrationally by:
a) honoring sunk costs
b) exhibiting menus dependence (decoy effect)
c) outweighing losses relative to gains
d) permitting arbitrary anchors and following various heuristics
• Note, however, in the face of these potential problems one does not
have to abandon rationality and opt for a behavioral approach. Later, we
will examine the notion of bounded rationality, which helps to account
for various "behavioral" patterns within the formal axiomatic approach
In the next lecture, we will examine rationality within the context of the
choice under uncertainty, which requires more assumptions and, thus, is
more restrictive than in the case of choice under certainty. The reading
assignment for this lecture is Chapter 2 and 3 from BE.