Lecture 1.

Mechanism Design
Fabio Castiglionesi
CentER, and Tilburg University

In this lecture we discuss mechanism design (also called ‘implementation’

theory). This general approach proves to be particularly useful in contract
theory. Indeed, a contract can be considered a mechanism that implements
some objective function of the agents involved in the contract. We introduce
the basic environment for mechanism design and we analyze how it can be
adapted to situations of hidden information and hidden action. The notions
of weak and (strong) implementation are introduced. We then prove the
important result called the Revelation Principle, which allows to restrict
considerably the range of mechanisms that are feasible. Implementation
in dominant strategies and the (strong) implementation problem are finally
analyzed.
Since mechanism design, and in general the course on asymmetric infor-
mation, implies a basic knowledge of game theory, in the Appendix of this
lecture we provide a quick review of game theoretic concepts that turn out
to be useful in various applications in financial economics. We define what
a game is and how game theorists predict the outcome of a game. We then
review a class of games particularly useful in finance: The so-called ‘signal-
ing games’. We finally provide a useful refinement of equilibrium concept in
signaling games.

1 Mechanism Design
If we have some theory which allows us to predict the outcome of a given
game, we may also pose the reverse question: If we want to obtain a certain
outcome, how should we design the rules of the game? This is what the mech-
anism design literature studies. Formally, the problem can be represented as

1
follows.
Consider a society composed of a set I of members, with i denoting the
generic member. There is a set of possible states of nature Θ, and each agent
is endowed with a partition PΘi of Θ. Furthermore, each agent i can take an
action ai in some set Ai . We denote by A = Πi∈I Ai the set of all possible
action profiles, with generic element a = {ai }i∈I . Action profiles may be
partially observable, and we will assume that each agent is endowed with a
partition PAi of A.
There is a set D of decisions that the society can take. A collective de-
cision rule (sometimes called a social welfare correspondence) is a mapping
from the possible states of the world and action profiles to (subsets of) de-
cisions, that is d : Θ × A → D. In other words, the society would like,
for each state of the world θ, and action profile a to have the decision d be
implemented.
Each agent is endowed with a utility function ui : Θ × D × A → , with
ui (θ, d, a) denoting the utility when the state of the world is θ, the decision
is d and the action profile is a.
Special cases of this general problem can be considered according to the
following criteria: whether the private information bears on who the agent
is (or what his characteristics are) or on what the agent does. The former
are called hidden information models,1 the latter are called hidden action (or
moral hazard) models.

1.1 Hidden Information

In this case we assume A to be singleton, so that no action is to be taken,
and the state of the world is private information of the agents. If there was
a benevolent social planner who can observe θ and choose the decision then
no problem would arise. However, in many situations such a planner does
not exist. The objective of the theory of mechanism design is to explore the
means to implementing a given allocation when the information is dispersed
in the economy. The information is only known to the agents, and in order
to exploit such information it is necessary to elicit truthfully the information
1
Among hidden information models we can distinguish models in which the initiative
belongs to the uninformed party (adverse selection models) or to the informed party (sig-
naling models).

2
from the agents, and then take the appropriate decision. In this case we can
write utilities as ui (θ, d), and a collective decision rule as d (θ) ⊂ D.

1.2 Hidden Actions

Assume now that the set Θ is singleton but there is at least one agent i ∈ I
such that Ai is not singleton. We can now write utility functions as ui (a, d).
In the standard moral hazard problem, a principal hires a team of agents
I = {1, . . . , n}. Each agent takes an action ai ∈ Ai , and the vector of actions
a = (a1 , . . . , an ) determines an outcome π (a, ν), where ν is a random variable
on which no participant (principal or agents) has information.
The principal wants a certain action profile a∗ = (a∗1 , . . . , a∗n ) to be taken.
Actions are unobservable by the principal, but agents can be rewarded with
compensation schemes (w1 (π) , . . . , wn (π)). Therefore, the set of decisions is
the set of outcome-contingent compensation rules, and the utility functions
depend on the action profile and on the monetary transfer, ui (a, m). Assum-
ing von Neumann-Morgenstern preferences, we can extend preferences over
compensation rules by setting ui (a, wi ) = E [ui (a, wi (π (a, ν)))].
In principle, a collective decision rule should specify a selection of the
compensation scheme for each possible action profile. However, since the
principal is interested in having a particular action profile a∗ enacted, we are
mostly interested in the compensation scheme to be selected when the action
profile a∗ is taken. We are relatively free to specify the collective decision
rule for a = a∗ as it turns out to be more convenient.
In the pure moral hazard model, agents only observe their own action,
so that PAi (a) = ai × A−i for each a and i, and their utility only depends
on their own action. Furthermore the sets Ai ’s are ordered sets, and in
most applications they are interpreted as efforts. Agents prefer less effort to
more. On the other hand, higher effort is preferred by the principal because
it induces higher values of π (say, profit of the firm). The literature has
focused on the case in which the principal is risk-neutral and the agents have
separable utility functions and are risk averse. In this case, an interesting
collective decision rule is the one in which the action profile a∗ is the one
that maximizes social surplus, and agents are rewarded with a fixed salary.
In most cases this decision rule cannot be implemented, so that the standard
trade-off between risk-sharing and incentive provision appears.
Incentive problems can be somewhat alleviated if the actions taken by

3
the agents can be (at least partially) observed by other players. One ex-
treme case is the one in which all agents observe perfectly the action profile,
PAi (a) = {a} for each i and a. In principle, in this case there is no private in-
formation on actions, and therefore the trade-off between incentive provision
and risk-sharing disappears. However, nontrivial problems of multiplicity of
equilibrium may arise.

2 The Revelation Principle with Hidden In-

formation
As anticipated, the mechanism design and the implementation literature are
concerned with the following question: Is there a way in which I can elicit
the information and the actions desired from the agents so that at each state
of the world the desired collective decision rule is enacted?
In order to answer this question we have to set up some notation and
clarify a few concepts. First, let us define a mechanism Γ = (M, g) as a
collection of message spaces M = (M 1 , . . . , M n ) and an outcome function
g : M → D. The interpretation is the following. M i is the message space of
agent i, describing the actions that she can take. After all agents have sent
their messages, a message profile m = (m1 , . . . , mn ) is determined. Depend-
ing on the message profile m, a certain decision g (m) is chosen. When we
add to this picture the utility functions of the agents and their information,
we obtain a game. Then we can use interchangeably the terms ‘mechanism’
and ‘game’.
We will have some solution concept S (like Nash, subgame perfect, Bayesian,
perfect Bayesian equilibrium) which guides us in predicting the outcome of
the game. Given a solution concept S, let us call S ((M, g) , θ) ⊂ D the set
of outcomes predicted by S when a mechanism (M, g) is offered to the agents
and θ is the state of the world.
Definition. A collective decision rule d (θ) is weakly implemented by the
mechanism (M, g) under solution concept S if d (θ) ⊂ S ((M, g) , θ) for
each θ.
Therefore, d (θ) is weakly implemented when each decision in d (θ) can
be obtained as the outcome of the mechanism when the state of the world is
θ.

4
 For example, assume that (M, g) induces a game of incomplete informa-
tion among players, and that the solution notion we adopt is that of Bayesian
equilibrium. Assume further that the collective decision rule is a function,
so that d (θ) is a singleton for each θ. Then we say that the mechanism
(M, g) realizes d (θ) if there is a Bayesian equilibrium σ : Θ → M such that
g (σ (θ)) = d (θ) for each θ.
The term ‘weak implementation’ refers to the fact that we do not insist
on the mechanism having only outcomes which we deem desirable (that is,
belonging to d (θ)). In fact, we only require that elements in d (θ) be among
the possible outcomes of the mechanism, leaving open the possibility of other
outcomes of the mechanism which may not be desirable. Of course, we
may want to look at a stronger form of implementation, requiring that only
desirable outcomes should occur. This is what motivates the next definition.

Definition. A collective decision rule d (θ) is (strongly) implemented by the

mechanism (M, g) under solution concept S if d (θ) = S ((M, g) , θ) for
each θ.

In this case we require that an outcome can result from the mechanism
if and only if it is in d (θ). The ‘only if’ part is what differentiates the
notion of (strong) implementation from the one of weak implementation. The
difference with weak implementation is that with (strong) implementation we
want all the equilibria of the mechanism Γ to deliver d (θ), rather than just
some (or one) of the equilibria of the mechanism to deliver d (θ). However,
there are significant differences between the set of collective decision rules
which are weakly implementable and the set of collective decision rules which
are implementable. We will only focus our attention on weak implementation.

2.1 Weak Implementation

When is it the case that d(θ) can be weakly implemented? The problem
appears very complex, since in principle we should look at the set of all
possible mechanisms Γ = (M, g), which appears to be quite intractable from
the mathematical point of view. Fortunately, an important result known
as the Revelation Principle allows us to reduce dramatically the space of
relevant mechanisms. In fact, the Revelation Principle tells us that the only
mechanism we have to worry about is the so-called direct mechanism.

5
 The direct mechanism is a mechanism in which each agent has a message
space M i = {PΩi (θ)}θ∈Θ , that is each agent simply announces the information
possessed. The planner collects all the information issued by the agents and
determines
  the state of the world θ
= ∩i∈I PΘi (θ). Then, the decision selected
is d θ
. The Revelation Principle tells us that, only if truthtelling is an
equilibrium in the direct mechanism, it is possible to weakly-implement the
decision rule d(θ). In a truthful mechanism each agent finds it optimal to
announce the true value of her information (or ‘type’).
Summing up, in the case of weak implementation (in dominant strategies
or in Bayesian equilibrium), the revelation principle allows us to restrict
attention to a particular class of mechanisms, the class of direct (truthful)
mechanisms.
The Revelation Principle. Suppose that there is a mechanism Γ = (M, g)
which weakly implements d(θ) in Bayesian equilibrium. Then d(θ) is
also implemented by the direct mechanism.
Proof. Since Γ = (M, g) weakly implements d(θ), the mechanism Γ has a
Bayesian equilibrium σ such that g (σ (θ)) = d (θ) for each θ. Define now the
new mechanism Γ = M , g

such that M i = {P i (θ)}
θ∈Θ and g = g(σ(.)) =
Θ

i = {P i (θ)}
d(.). Since M 
Θ θ∈Θ the mechanism Γ is a direct mechanism. We
now show that truthtelling is an equilibrium of the direct mechanism.
Suppose not. Then there is at least one
  agent i and state ω

such that the
i i

agent prefers to announce PΘ θ = PΘ θ . Let:
          
E ui g σ i PΘi θ , σ −i PΘ−i (θ) , θ  PΘi θ

be the expected
  utility of agent i when the other
 agents
 tell the truth, infor-
i
i
mation PΘ θ is observed and information PΘ θ is announced by agent i.
 
Then, since PΘi θ is a profitable deviation from truthtelling:
          
E ui g σ i PΘi θ , σ −i PΘ−i (θ) , θ  PΘi θ
> (1)
          
E ui g σ i PΘi θ
, σ −i PΘ−i (θ) , θ  PΘi θ

  
Consider now the original mechanism Γ, and let mi = σ i PΘi θ i =
and m
  
σi PΘi θ
. Then, inequality [1] implies:
       
E ui g mi , σ −i PΘ−i (θ) , θ  PΘi θ
>

6
        

E ui g m
i , σ −i PΘ−i (θ) , θ  PΘi θ

The last inequality says
 that
 agent i prefers to use the message intended to
reveal information PΘi θ under the strategy profile σ (even if she observes
   
PΘi θ
). In other words, when the state of the world is given by PΘi θ
,
agent
 i prefer to use the message intended to reveal the state of the world
PΘi θ under the strategy profile σ (i.e., agent i finds convenient to deviate).
This contradicts the fact that σ is an equilibrium in the original mechanism
(M, g).
Remark. A version of the Revelation Principle also holds for dominant
strategy implementation, that is when we predict that agents only select
dominant strategies (the proof is easily adapted from the one above).
The revelation principle can be interpreted as follows. Each agent dele-
gates her choice of a strategy to a planner. The planner asks to the player
what her type is, and with this information chooses a strategy in order to
maximize the player’s playoff. The revelation principle states that no agent
has incentive to fool the mediator by telling false information, because this
would result in the mediator picking up not the best strategy.
The revelation principle allows us to focus our attention on the properties
of the collective decision rule d, rather than on the institutional details which
allow the implementation of the rule. It says that, as long as we are satisfied
with weak implementation, either the simplest institution (i.e. the direct
mechanism) works, or no institution works. No complicated ways of eliciting
the agents’ information need to be taken into account.2
The following property is crucial for weak-implementation.

Definition. A decision rule d satisfies incentive compatibility if

        
E ui d PΘi θ
, PΘ−i (θ) , θ  PΘi θ
≥
        
E ui d PΘi θ , PΘ−i (θ) , θ  PΘi θ

 

θ .
for each agent i and each pair θ,
2
An issue we are going to ignore is that information itself may be difficult to commu-
nicate, for example because the elements PΩi (ω) are complicated objects which are hard
to describe with precision.

7
 A decision rule satisfies incentive compatibility (IC) if at each state of the
world, each agent prefers to announce the truth when the other agents are
announcing the truth.3 Then we immediately have the following result.

Theorem. A collective decision rule d is weakly implementable if and only

if it is incentive compatible.

Proof. For the ‘if’ part (sufficient condition), simply observe that when d
satisfies incentive compatibility, the direct mechanism weakly implements d.
For the ‘only if’ part (necessary condition) we invoke the revelation princi-
ple. If d does not satisfy incentive compatibility then truthtelling is not an
equilibrium of the direct mechanism. But then, by the revelation principle,
no other mechanism can weakly implement d.
Then, in case of weak implementation (i.e., it is enough to have a mech-
anism in which one of the equilibria is the desired outcome), the incentive
compatibility condition is a necessary and sufficient condition for implemen-
tation. All we have to do is to check whether truthtelling is an equilibrium in
the direct mechanism. If this is not the case, then the outcome is not weakly
implementable.
Incentive compatibility does not assure instead (strong) implementation.
The difference on the conditions between weak implementation and (strong)
implementation is due to the fact that the truthful strategy may not be
the only equilibrium strategy in the direct mechanism, even if the incentive
compatibility constraint is satisfied. In other words, there can be several
equilibria in the direct mechanism: In one of them players are using truth-
ful strategies, but in the other equilibria the players are using untruthful
strategies. If such untruthful equilibria exist, unless there is a good reason
to discard all but the truthful equilibrium, truthful implementation does not
imply (strong) implementation.
In principle, there are no good reasons to consider only the truthful equi-
librium in a direct mechanism. However, this is the attitude that has pre-
3
Notice that we have written the function d as a function of reported information
rather than as a function of the state of the world ω. This is necessary in order to take
care of situations in which the information reported by the agents is incompatible, that
is the intersection of the reported sets PΩi is empty. When we describe uncertainty with
the ‘type’ structure, this is equivalent to agents announcing a profile of types having
probability zero. In order to ensure incentive compatibility, the function d has to specify
the decision to be taken in those cases.

8
vailed in the literature because the results are clearer. If we want to elimi-
nate all the equilibria that are not desired then the direct mechanism is not
enough. More complex mechanisms are needed.

3 Dominant Strategies
We would like to be able to implement a collective decision rule by imposing
the weakest behavioral requirements on the agents. One weak requirement is
that agents only pick dominant strategies. This implies that we do not have
to worry about the conjectures that each agent may hold about other agents’
behavior. So why do we choose to use Bayesian equilibrium as solution
concept instead of dominant strategy to analyze weak implementation?
The reason is that weakly implementing a decision rule in dominant
strategies may be very hard in some contexts. An important result due
to Gibbard and Sattherwaite states that if a decision rule is implementable
in dominant strategies and the agents may rank elements in D in all possible
ways, then the only decision rules which are weakly implementable in domi-
nant strategies are the ones which are dictatorial, that is rules which always
select the alternative preferred by a particular individual (the dictator).
There are two ways out of this negative result. The first is to use a
different solution concept to predict the outcome of the mechanisms proposed
to the agents (obvious candidates are less restrictive solution concepts such
as Bayesian equilibrium as done before). The other possibility is to retain the
requirement that agents play dominant strategies, but limit the variability of
the environment. We will focus on this second alternative in the next section.

3.1 Clarke-Groves Mechanisms

When the decision to be taken involves monetary transfers, it is pretty natural
to assume that each agent should like more money to less; we can therefore
rule out preferences which do not exhibit monotonicity in the amount of
money assigned to the agent. This apparently innocuous assumption is of
great help in circumventing the negative results of the Gibbard-Sattherwaite
theorem.
Consider the traditional problem of producing a public good. Let X =
{0, 1} denote the production decision, and M i a subset of , with mi ∈ M i

9
denoting the transfer to agent i. Then D = X × M 1 × . . . M n .
The utility function of agent i is ui (d, mi ) = xv i + mi . The cost of the
project is C, and a typical state of the world is a vector θ = (v 1 , . . . , v n ).
We would like to weakly implement the efficient rule, stating that x (θ) = 1

iff ni=1 vi ≥ C. The problem is that the agents may have an incentive to
understate or overstate their valuations, in order to increase or decrease the
probability of producing the public good. Therefore, in order to implement
the rule we need to devise a system of transfers m1 (θ) , . . . , mn (θ) such that
every agent has an incentive to tell the truth. Notice that here we depart
somewhat from the framework spelled out in the previous section. In princi-
ple, a decision rule describes both the production and the transfer decision.
Here we focus only on the production decision, permitting any kind of trans-
fers. This is important because, as we will shortly see, implementation in
dominant strategies becomes impossible as soon as we impose some reason-
able restriction on the class of transfers allowed.
At any rate, here is a mechanism that implements in dominant strategies
the efficient rule. Let v
i be the announcement by agent i. Then:
  n
  

j − C + h (v
j=i v
−i ) if
j ≥ C
j=1 v
mi v
1 , . . . , v
n = 
 n
h (v
−i ) if j=1 v
j < C

Under this transfer scheme, for any vector of announcements v
−i by the other
players the agent solves:
 
    
max x v
i , v
−i v i + v
j − C  + h v
−i


vi j=i

It is clear that by announcing the truth v
i = vi the agent maximizes the
objective function. We therefore conclude that truthtelling is implemented
in dominated strategies.
The main problem with the Clarke-Groves scheme is that there may be
no way to make sure that the budget balances. In particular, notice that we
can exploit the quasi-linearity assumption and the lack of IR constraints to
make sure that the sum of transfers is always negative, but we cannot make
sure that it is zero. A solution to the problem can be found if we give up
dominant strategies and accept to use Bayesian equilibrium as the solution
notion.

10
4 Appendix: Review of Game Theory
This appendix does not represent an exhaustive treaty of game theory. A
complete and in deep treaty is given by Fudenberg and Tirole (1991). A more
introductory level of game theory is provided by Gibbons (1992). Salanié’s
appendix (1997) provides a summary of the main game theoretic concepts
used in contract theory.

4.1 Definitions of Games and the Notions of Equilib-

rium
We will present four classes of games: i) normal form (static) games of com-
plete information; ii) extensive form (dynamic) games of complete informa-
tion; iii) normal form games of incomplete information, and iv) extensive
form games of incomplete information. A game has complete information
if the player’s utility (or payoff) functions are common knowledge. Games
with incomplete information (also called Bayesian games) are characterized
by the fact that one player does not know another player’s payoff function.
Game theory tries to predict the behavior of agents when they face strate-
gic situations. The main tool used is the notion of equilibrium. Generally
speaking, an equilibrium is a situation in which each agent follows a strategy
that maximizes her utility given the strategies of other players. The exact
meaning of this statement varies depending on whether we consider normal
form games or extensive form games. Consequently, corresponding to the
previous four classes of games we have four notions of equilibrium: i) Nash
equilibrium; ii) subgame-perfect equilibrium; iii) Bayesian equilibrium, and
iv) perfect Bayesian equilibrium.
These equilibrium concepts are closely related. As we consider richer
games, we progressively strengthen the equilibrium concept, in order to rule
out implausible equilibria in the richer games that would survive if we applied
equilibrium concepts suitable for simpler games. In each case, the stronger
equilibrium concept differs from the weaker concept only for the richer games,
not for the simpler games.

11
4.1.1 Normal Form Games with Complete Information
These class of games are of the following simple form: first the players simul-
taneously choose actions; then the players receive payoffs that depend on the
combination of actions just chosen. Each player’s payoff function, i.e. the
function that determines the player’s payoff from the combination of actions
chosen by the players, is common knowledge among players.
Let I = {1, . . . , n} be a set of players. An n−person normal form game
of complete information G is an array:

G = {A1 , . . . , An , u1 , . . . un }

where:

• Ai is the set of actions available to agent i, with typical element ai ∈ Ai ;

• ui is the utility (or payoff) function of agent i. It is a map ui : A → ,

where A = ×nj=1 Aj is the Cartesian product of the action spaces, with
typical element a = (a1 , . . . an ). It is assumed that ui satisfies the Von
Neumann—Morgenstern (expected utility) assumptions.

We will frequently use the notation A−i = ×j=i Aj to denote the set of
possible action profiles that can be undertaken by agents other than i. The
typical element of A−i will be denoted a−i = (a1 , . . . ai−1 , ai+1 , . . . an ), and
when we want to emphasize in a given action profile a ∈ A the action of
agent i we write a = (ai , a−i ).
In the case of normal form games, a strategy σi by agent i can be described
as a probability distribution on the set Ai . In general, given any set X we
will denote by ∆X the set of probability distributions over X. Thus, a
strategy is an element of ∆Ai . Given any action ai , the symbol σi (ai ) denotes
the probability that agent i will choose action ai . More generally, given
any subset Ai ⊂ Ai , the probability of taking an action in the set Ai is

σi (Ai ) = ai ∈A
i σi (ai ). A strategy by agent i is called pure if the support
of σi is a singleton set (i.e., {ai }); it is called mixed otherwise. Since in
a normal form game it is assumed that agents take their action in isolation
and independently, a given collection of strategies (σ1 , . . . , σn ) can be used to
define a probability distribution over A. We simply have that the probability
of an action profile a = (a1 , ...an ) is given by σ(a) = σ1 (a1 )·σ2 (a2 )·. . .·σn (an ).

12
 If agent i chooses action ai and the other players are using the strategies
σ−i then the expected utility of agent i is:

Ui (ai , σ−i ) = ui (ai , a−i ) σ−i (a−i )
a−i ∈A−i

A Nash equilibrium is a collection of strategies σ ∗ = (σ1∗ , . . . , σn∗ ) such that

for each agent i, each action in the support of σi∗ maximizes the expected
∗
utility when other agents adopt the strategy σ−i . More precisely, for each
∗
ai ∈supp σi we have:
   
∗
Ui ai , σ−i ≥ Ui ai , σ−i
∗
for each ai ∈ Ai

A Nash equilibrium is said to be in pure strategies if each agent adopts a

pure strategy. This implies that the support of σ ∗ is singleton.
One way to motivate the definition of Nash equilibrium goes as follows.
Suppose that game theory makes a unique prediction about the strategy each
player will choose. In order for this prediction to be correct, it is necessary
that each player be willing to choose the strategy predicted by the theory.
Thus, each player’s predicted strategy must be that player’s best response
to the predicted strategies of the other players. Such a prediction could be
called strategically stable, because no single player wants to deviate from her
predicted strategy. A related motivation for Nash equilibrium involves the
idea of convention: if a convention is to develop how to play a given game
then the strategies prescribed by the convention must be a Nash equilibrium,
else at least one player will refuse the convention.

4.1.2 Extensive Form Games with Complete Information

In this class of games players choose actions sequentially (and not simul-
taneously). For this reason these games are also called dynamic games.
Here dynamic games with complete information are taken into considera-
tion. However, in this class of games we can have perfect information, i.e. at
each move in the game the player with the move knows the full history of
the play of the game thus far, or imperfect information, i.e. at some move
the player with the move does not know the history of the game.
A game tree T is a collection {N, ≺}, where N is the set of nodes and ≺
is a partial order of set N . Given two elements n and n in N we say that n

13
precedes n if n ≺ n . Equivalently, we say that n is a successor of n. If a
node precedes all other nodes then that node is called the origin (or root) of
the game. If a node has no successors then it is called a terminal node. We
will call Z the set of terminal nodes. A finite n−person extensive form game
of complete information Γ is an array:

Γ = {{N, ≺} , ι, B1 , . . . , Bn , C1 , . . . Cn , u1 , . . . un }

where:

• ι : N\Z → {1, . . . , n} is a mapping assigning a player to each node (by

convention, terminal nodes are excluded);

• Let N i = { n| ι (n) = i} be the subset of nodes in which agent i de-

cides. Then Bi is a partition of N i , that is B i is a collection of subsets
1 mi mi k i k k

{b
 i , . .
., bi } such that ∪k=1 bi = N and bi ∩ bi = ∅ for each pair
bki , bki . Each element of Bi is called an information set. For a given
node n, we denote by b (n) the information set to which n belongs.
k
• Ci is a mapping
 associating to each information set bi of agent i an
action set Ci bki .

• ui : Z →  is the utility function of player i.

The semantic convention has prevailed that an extensive form game is said
to be of perfect information if each information set is singleton, otherwise it
is called of imperfect information. A subgame of an extensive form game Γ is
obtained when a certain node n0 can be used as the root of a tree. We don’t
give here the formal definition. Notice that, by definition, Γ is a subgame of
itself. A subgame Γ excluding some nodes of Γ is called a proper subgame.
The definition of strategy is slightly more complicated for extensive form
games. Remember that Bi = {b1i , . . . , bm i } is the collection of information
i

sets of agent i, and Ci bki is the set of possible actions at information set
bki . Then
 a strategy  for agent i is a collection of probability distributions
σi = σb1i , . . . , σbm i , where each probability distribution σbk is defined on
  i i

the set Ci bki . For any subgame Γ of the extensive form game Γ we can
define the restriction of the strategy of player i to the subgame Γ simply as

14
the collection of probability distributions σbki such that the information set
bki belongs to the subgame. The restriction is denoted σi |Γ
.
When a strategy profile σ = (σ1 , . . . , σn ) is given, we obtain a probability
distribution over the set of final nodes Z. Similarly, for any subgame Γ , we
can define the restricted strategy profile σ|Γ
= (σ1 |Γ
, . . . , σn |Γ
), and such
restricted strategy profile delivers a probability distribution over Z  , the set
of terminal nodes of Γ .
A subgame perfect equilibrium is a collection of strategies σ ∗ = (σ1∗ , . . . , σn∗ )
such that for each subgame Γ the restriction σ ∗ |Γ
is a Nash equilibrium. No-
tice that the definition implies that a subgame perfect equilibrium is a Nash
equilibrium, since the game Γ is a subgame of itself. Therefore, the set of
subgame perfect equilibria of a given game is contained in the set of Nash
equilibria of that game. The two sets coincide if the game has no proper
subgames. There are however games having Nash equilibria which are not
subgame perfect equilibria. The reason is that the concept of Nash equilib-
rium does not put any restriction on strategies adopted on subgames which
are not reached with positive probability, while subgame perfection insists
on rational behavior on unreached subgames as well.

4.1.3 Normal Form Games with Incomplete Information

Recall that in a Bayesian game at least one player is uncertain about another
player’s payoff function. A common example of a static game of incomplete
information is a sealed-bid auction: each bidder knows her own valuation for
the good being sold but does not know any other bidder’s valuation. Bids
are submitted in sealed envelops such that players’ moves can be considered
simultaneous. However, most economically interesting Bayesian games are
dynamic, i.e. represented in extensive form (see next section).
An n−person normal form game of incomplete information Gµ , otherwise
called a Bayesian game, is an array:

Gµ = {Θ1 , . . . , Θn , µ, A1 , . . . An , u1 , . . . , un }

where:

• Θi is the set of types of agent i, with typical element θi . It represent

the information available to agent i;

15
 • µ is a probability distribution over Θ = ×ni=1 Θi . The set Θ is called
the set of the states of the world, and it has typical element θ. The
notations Θ−i and θ−i have obvious meaning.

• Ai is the set of actions available to agent i;

• ui : A × Θ →  is the utility function of agent i. Thus, ui (a, θ) is the

utility attained by agent i when the realized action profile is a and the
state of the world is θ.4

Following Harsanyi (1967), it is commonly assumed that the timing of

a static Bayesian game is as follows. First, nature draws a type vector θ =
(θ1 , ..., θn ), where θi is drawn from the set of possible types Θi . Second, nature
reveals θi only to player i. Third, players simultaneously choose actions,
player i choosing a from the feasible set Ai . Finally, payoffs ui (a, θ) are
received. Note that, since nature reveals player’s i type only to player i,
player j does not know the complete history of the game when actions are
chosen. This means that, by introducing the fictional move by nature in
the first two steps, we have described a game of incomplete information as a
game of imperfect information.
In games of incomplete information, a strategy for agent i is defined as
a mapping σi : Θi → ∆Ai . A strategy describes the probability distribu-
tion over actions adopted by each possible type of agent i. We will use the
symbol σi ( ai | θi ) to denote the probability that type θi of agent i takes ac-
tion ai . The ex ante probability that agent i will chose action ai is given by

θi ∈Θi σi ( ai | θi ) µ (θi ), where µ (θi ) is the probability that type θi occurs. A
strategy profile σ = (σ1 , . . . σn ) defines a probability distribution over Θ × A.
Alternatively, we can say that the strategy profile defines a conditional prob-
ability distribution σ ( ·| θ) over A for each state of the world θ ∈ Θ. Given a
strategy profile σ, the expected utility of type θi of agent i taking action ai
4
In principle, the set of actions Ai available to agent i may depend on her type. For
example, θi may be the production capacity of a firm, and Ai the possible quantities
produced. High quantities may be available only to high capacity firms. However, an
equivalent way to treat the situation is just to assign a utility of −∞ to an agent who
takes an action which is not allowed, no matter what other agents do. If this trick is
allowed then assuming that Ai is the same for all types is without loss of generality.

16
is given by:
 
 
Ui (ai , σ−i | θi ) =  ui ((ai , a−i ) , (θi , θ−i )) σ−i ( a−i | θ−i ) µ (θ−i | θi )
θ−i ∈Θ−i a−i ∈A−i

where σ−i (a−i | θ−i ) is the probability that action profile a−i is taken by agents
other than i when their type profile is θ−i , and µ ( θ−i | θi ) is the conditional
probability of θ−i given θi . In order to compute µ (θ−i | θi ), it is assumed that
it is common knowlwdge that nature draws the type vector θ = (θ1 , ..., θn )
according to the prior distribution µ (θ). Then when nature reveals θi to
player i, she can compute the belief µ ( θ−i | θi ) using Bayes’ rule:

µ (θ−i , θi ) µ (θ−i , θi )
µ ( θ−i | θi ) = = .
µ (θi ) θ−i ∈Θ−i µ (θ−i , θi )

A Bayesian equilibrium is a strategy profile σ ∗ = (σ1∗ , . . . σn∗ ) such that for

each agent i, for each type θi of agent i, and for each ai ∈suppσi∗ (·| θi ) :
     
∗  ∗ 
Ui ai , σ−i  θi ≥ Ui ai , σ−i  θi for each ai ∈ Ai .

That is, no player wants to change her startegy, even if the change involves
only one action by one type.

4.1.4 Extensive Form Games with Incomplete Information

The definition of n-person extensive form game of incomplete information Γµ ,
can be obtained immediately from the definitions before: it is an extensive
game form Γ plus a collection of sets Θi and a probability distribution µ,
with utility functions defined over Z × Θ.
For extensive form games of incomplete information (or, complete but
imperfect information) the notion of perfect Bayesian equilibrium is used.
We will consider here games of complete but imperfect information, the
introduction of incomplete information is straightforward but notationally
cumbersome.
To introduce the notion, first define a probability assessment Ψ for a given
game Γ as an assignment to each information set of a probability distribution
over the nodes of the set. We will denote by µb the probability distribution for
information set b, so that a probability assessment is a collection Ψ = {µb }.

17
Given a strategy profile σ we say that a probability assessment Ψ is consistent
with σ if, for each information set b reached with positive probability when
σ is used, the probability µb is obtained using the Bayes’ rule. At each
information set bki , the agent who has to move can compute her expected
utility for each possible action ai ∈ C bki . In order to do this, she needs a
probability distribution over the nodes of bki and a probability distribution
over the terminal nodes Z. The first element is provided by the probability
distribution µbki , while the second element can be computed using the strategy
profile σ. For each node n ∈ bki define π (z| n, ai , σ−i ) the probability of
reaching terminal node z starting from node n given that agent i takes action
ai and the other agents use the strategy profile σ−i . Then an action ai gives
to a player i at information set bki and holding a probability distribution µbki
the following expected utility:
 
   
Ui ai , σ−i | bki , µbki = ui (z) π (z| n, ai , σ−i ) µbki (n)
n∈bki z∈Z

where µbki (n) is the probability assigned to node n.

A perfect Bayesian equilibrium is a pair (σ ∗ , Ψ) such that:

• Each agent maximizes expected utility at each information set, where

expected utility is computed using the probability distributions given
by Ψ. Formally, for each i, for each bki and for each ai ∈suppσbki we
have:
       
∗  k 
Ui ai , σ−i  bi , µbk ≥ Ui ai , σ−i
∗
 bki , µbk ∀ai ∈ C bki
i i

• The probability assessment Ψ is consistent with σ ∗ .

Notice that the only requirement on Ψ is that it be consistent with σ ∗ ,

that is that the Bayes’ rule be used whenever possible. This means that
in a perfect Bayesian equilibrium no restriction is imposed on information
sets which are not reached with positive probability. Various refinements of
this notion have been proposed. In general the main idea is to impose some
extra restrictions on the way in which beliefs can be formed, ruling out some
‘unreasonable’ beliefs on unreached information sets. We will discuss one of
these possible refinements in the next section.

18
4.2 Signaling Games
A class of extensive form games of incomplete information which has been
frequently used in finance, as well as in other fields, is that of signaling games
(also called ‘sender-receiver games’).
The simplest signaling game has the following structure:

• There are two players, the ‘sender’ and the ‘receiver’.

• The type of sender is not known to the receiver, while the type of the
receiver is common knowledge. Let Θ be the set of sender’s types, with
typical element θ, and µ the probability distribution on Θ.

• The sender takes an action a1 ∈ A1 . The action is observed by the

receiver. After observing a1 the receiver chooses an action a2 ∈ A2 .

• The final payoff to agent i, with i = 1, 2 is given by a function ui (a1 , a2 , θ).

This structure can be generalized to include multiple receivers. The cru-

cial characteristic of signaling games is that there is one party (the sender)
who has the private information, and another party (the receiver) who chooses
the final outcome. This implies that the receiver will try to use the move
by the sender to obtain information on her type. Here are some examples of
signaling games arising in finance:

• Dividend games. The sender is the management of the firm, the re-
ceivers are investors. The private information held by the management
is the value of the firm. The action taken by the management is the
determination of the level of dividends. The action taken by investors
is the provision of capital to the firm.

• Capital structure games. The situation is as before, but now the man-
agement decides the mix of debt and equity to be used to finance the
firm.

• Takeover games. The sender is a bidder who moves first, making an

offer for the firm. The private information is the value of the firm to
the bidder. Receivers are other potential bidders, who have to decide
whether or not to enter into the race.

19
A perfect Bayesian equilibrium in a signaling game is given by a strategy
profile σ ∗ = (σ1∗ , σ2∗ ) and a probability assessment Ψ = { µ ( ·| a1 )| a1 ∈ A1 }.
Here σ1∗ : Θ → ∆A1 gives the probability with which each action is chosen
by each type of player 1, σ2∗ : A1 → ∆A2 gives the probability with which
player 2 chooses each action given the move of player 1, and µ ( ·| a1 ) gives
the revised probability distribution on Θ given a1 . The probability µ ( ·| a1 )
is computed using the Bayes’ rule whenever possible, that is:

µ (θ) σ1∗ ( a1 | θ)
µ ( θ| a1 ) = 
θ
∈Θ µ (θ ) σ (a1 | θ )
 ∗ 

 
whenever θ
∈Θ µ (θ ) σ ∗ ( a1 | θ ) > 0. Notice here that θ
∈Θ µ (θ ) σ ∗ (a1 | θ  )
is the total probability that action a1 will be chosen by agent 1, while the
numerator is the joint probability that the type is θ and action a1 is selected.
There are two types of equilibria in signaling games: separating equilibria
and (partially) pooling equilibria. An equilibrium is separating if the action
chosen by the sender completely reveals her type. Formally:

suppσ1 ( ·| θ) ∩ suppσ1 ( ·| θ  ) = ∅ for each pair (θ, θ  )

In a partially pooling equilibrium, such a condition is not satisfied. This

means that there is at least one action a1 which is chosen with positive
probability by two different types θ and θ , so that observation of a1 leaves
the receiver uncertain about the type of the sender. An equilibrium is totally
pooling if suppσ1 ( ·| θ) is the same for each θ, so that no type can be excluded
after observing an action taken with positive probability in equilibrium.5

4.2.1 Refinement
Signaling games have usually a multiplicity of perfect Bayesian equilibria.
The main reason for this multiplicity is the relative freedom in specifying
beliefs on information sets which are reached out-of-equilibrium. Imposing
reasonable restrictions on out-of-equilibrium beliefs, can help in reducing the
number of equilibria. A classic example is the quiche-beer game, proposed
5
Notice however that even in a totally pooling equilibrium, the probability distribution
may change after a certain action a1 ∈suppσ1 is taken. This happens when different types
take the action with different probability.

20
by Cho and Kreps (1986).

What are the perfect Bayesian equilibria of this game? Let us indicate
with p the probability that the second player assigns to the weak type (tw )
after observing the action ‘quiche’, and with 1 − p the probability that player
2 assigns to the strong type (ts ) after observing the action ‘quiche’. Accord-
ingly, let us indicate with q the probability that the second player assigns to
the weak type after observing the action ‘beer’, and with 1−q the probability
that player 2 assigns to the strong type after observing the action ‘beer’.
First observe that there cannot be a separating equilibrium. In a sepa-
rating equilibrium the two types of agent 1 take different actions, and agent
2 knows the type of agent 1 when it is her turn to move. Thus agent 2 will
choose N (don’t duel) when the type is strong and F (duel) when the type
is weak. But this cannot be an equilibrium, since the weak type of 1 can
pretend to be strong (use the action of strong) and be better off.
There are two pooling perfect Bayesian equilibria:

[(Q, Q), (N, F ), p = 0.1, q ≥ 0.5],

[(B, B), (F, N), p ≥ 0.5, q = 0.1].

In the first pooling equilibrium, agent 1 always chooses Q irrespective of

the type, in the second pooling equilibrium B is chosen irrespective of the

21
type. In both equilibria, agent 2 chooses N after observing the equilibrium
move (i.e., after observing Q in the first case and B in the second case). This
action is sustained in both equilibrium by assigning a probability to the weak
type of p = 0.1 and q = 0.1, respectively. For example, in the first pooling
equilibrium, N is better than F (for agent 2) since with p = 0.1 the expected
payoff of N (i.e., 0.9) is greater than the expected payoff of F (i.e., 0.1).
If an out-of-equilibrium move is observed (i.e. B in the first equilibrium
and Q in the second), then agent 2 chooses F . The way in which it is
sustained, in the first pooling equilibrium, is by assigning a high probability
to the weak type when B is observed (i.e., q ≥ 0.5) so that agent 2 chooses F .
In the same way, in the second equilibrium it is sustained by assigning a high
probability to the weak type when Q is observed (i.e., p ≥ 0.5). For example,
in the first equilibrium when the out-of-equilibrium move B is observed, the F
is preferred to N if and only if q(1)+(1−q)(0) ≥ q(0)+(1−q)(1) =⇒ q ≥ 0.5.
The first equilibrium appears to be particularly non-intuitive. The reason
why this is non-intuitive is that weak is surely worse off by choosing B with
respect to the equilibrium; the payoff in equilibrium is 3, while at best she gets
2 after the deviation. Then the weak type cannot improve on the equilibrium
payoff choosing B rather than Q. The same reasoning does not apply to type
strong. In fact, the strong type could improve on the equilibrium payoff of
2, by receiving the payoff of 3. Hence, it seems reasonable to increase the
probability of strong when B is observed. In particular, it seems reasonable
to put q = 0 that is incompatible with the pooling equilibrium (Q, Q).
On the other hand, the beliefs in the second pooling equilibrium are not
suspicious. In fact, the weak type can be better off by choosing Q with respect
to the equilibrium (in equilibrium she gets 2, while she can get 3 after the
deviation). The strong type cannot improve choosing Q with respect to the
equilibrium (she gets 3, while at most she can get 2 after the deviation).
Hence, in this case it is reasonable to assign a high probability to the weak
type when the out-of-equilibrium move is observed (i.e., Q).
Cho and Kreps (1986) have proposed the following intuitive criterion to
reduce the set of equilibria in a signaling game.

Intuitive Criterion. If the information set following a1 is off-the-equilibrium

path and a1 is equilibrium dominated for type θi , then the Receiver’s
belief µ( θi | a1 ) should place zero probability on type θi . (This is possi-
ble provided a1 is not equilibrium-dominated for all the types in Θ).

22
It can be checked that in the quiche-beer game the equilibrium in which both
types choose Q does not survive the intuitive criterion. Consequently, the
set of perfect Bayesian equilibria of the game is made of only one equilibrium
once we impose the refinement proposed by Cho and Kreps.

References
[1] Cho, I.K. and D. Kreps (1987), ‘Signaling Games and Stable Equilibria’,
Quarterly Journal of Economics, 102: 179 -221.

[2] Fudenberg, D. and J. Tirole (1991), ‘Game Theory’, MIT press, Cam-
bridge MA.

[3] Gibbons, R. (1992), ‘Game theory for applied economists’, Princeton Uni-
versity press, Princeton.

[4] Harsanyi, J. (1967), ‘Games with Incomplete Information Played by

Bayesian Players Parts I, II and III’, Management Science, 14: 159-182,
320-334, 486-502.

[5] Salanié, B. (1997), ‘The Theory of Contracts: A Primer’, MIT press,

Cambridge MA.

23