SMU H3 Notes Game TheoryGamesSMU H3

Signalling Entry Game

Game theory analysis: Signalling Entry Game.


Setup

Definition:

Signalling Entry Game

  • Players: An incumbent firm (P1), and an entrant firm (P2).
  • Strategies:
    • Nature chooses the incumbent’s marginal cost mc1{5,15}mc_1\in\{5,15\} with probabilities α\alpha and 1α1-\alpha.
    • The incumbent chooses p1{15,20}p_1 \in \{15, 20\}.
    • The entrant observes the price and chooses In or Out.
  • Rules:
    • The incumbent has marginal cost 55 or 1515 and fixed cost 00.
    • The entrant has marginal cost 1010 and fixed cost 4040.
    • If entry does not occur, the incumbent firm has the full market quantity.
p=25qp = 25 - q
  • If entry occurs, firms play a Cournot duopoly.
p=25(q1+q2)p = 25 - (q_1 + q_2)

Game Tree

diagram

Payoff Details

Round 1

π1=pQc1q\pi_1 = pQ - c_1q π1=(25Q)Q5Q=(20Q)Q\pi_1 = (25-Q)Q-5Q=(20-Q)Q Q=10,p=15,π1=100Q=10,\qquad p=15,\qquad \pi_1=100 π1=pq1c1q1\pi_1 = pq_1 - c_1q_1 π1=15q115q1=0\pi_1=15q_1-15q_1=0 q1=0,p=16,π1=0q_1=0,\qquad p=16,\qquad \pi_1=0 π1=pq1c1q1\pi_1 = pq_1 - c_1q_1 π1=20q115q1=(10q1)q1\pi_1=20q_1-15q_1=(10-q_1)q_1 q1=5,p=20,π1=25q_1=5,\qquad p=20,\qquad \pi_1=25

Round 2

p=25(q1+q2)p = 25 - (q_1+q_2)

For P1 (incumbent):

π1=pq1c1q1\pi_1 = pq_1 - c_1q_1 π1=q1(25q1q2c1)=q1(20q1q2)\pi_1 = q_1(25 - q_1 - q_2 - c_1) = q_1(20 - q_1 - q_2)

Solving for FOC:

q1E(π1)=202q1q2=0\frac{\partial}{\partial q_1} E(\pi_1) = 20 - 2q_1 - q_2 = 0 q1=1012q2q_1 = 10 - \frac{1}{2}q_2

For P2 (entrant):

π2=pq2mc2q2fc2\pi_2 = pq_2 - mc_2q_2 - fc_2 π2=q2(25q1q2mc2)fc2=q2(15q1q2)40\pi_2 = q_2(25 - q_1 - q_2 - mc_2) - fc_2 = q_2(15 - q_1 - q_2) - 40

Solving for FOC

q2E(π2)=15q12q2=0\frac{\partial}{\partial q_2} E(\pi_2) = 15 - q_1 - 2q_2 = 0 q2=15212q1q_2 = \frac{15}{2} - \frac{1}{2}q_1

Equating the two quantities

q1=2538,q2=1033,p=40313q_1 = \frac{25}{3} \approx 8, \qquad q_2 = \frac{10}{3} \approx 3, \qquad p = \frac{40}{3} \approx 13 π1=625969,π2=260929\pi_1 = \frac{625}{9} \approx 69, \qquad \pi_2 = -\frac{260}{9} \approx -29 p=25(q1+q2)p = 25 - (q_1+q_2)

For P1 (incumbent):

π1=pq1c1q1\pi_1 = pq_1 - c_1q_1 π1=q1(25q1q2c1)=q1(10q1q2)\pi_1 = q_1(25 - q_1 - q_2 - c_1) = q_1(10 - q_1 - q_2)

Solving for FOC:

q1E(π1)=102q1q2=0q1=512q2\frac{\partial}{\partial q_1} E(\pi_1) = 10 - 2q_1 - q_2 = 0 \\ q_1 = 5 - \frac{1}{2}q_2

For P2 (entrant):

π2=pq2mc2q2fc2\pi_2 = pq_2 - mc_2q_2 - fc_2 π2=q2(25q1q2mc2)fc2=q2(15q1q2)40\pi_2 = q_2(25 - q_1 - q_2 - mc_2) - fc_2 = q_2(15 - q_1 - q_2) - 40

Solving for FOC:

q2E(π2)=15q12q2=0q2=15212q1\frac{\partial}{\partial q_2} E(\pi_2) = 15 - q_1 - 2q_2 = 0 \\ q_2 = \frac{15}{2} - \frac{1}{2}q_1

Equating the two quantities:

q1=532,q2=2037,p=50317q_1 = \frac{5}{3} \approx 2, \qquad q_2 = \frac{20}{3} \approx 7, \qquad p = \frac{50}{3} \approx 17 π1=2593,π2=4095\pi_1 = \frac{25}{9} \approx 3, \qquad \pi_2 = \frac{40}{9} \approx 5

Derivation (Best Response Analysis)

Separating Equilibrium

(p(5),p(15))(15,20)(p(5),p(15)) \mapsto (15,20) p=15cI=5,p=20cI=15p=15 \Rightarrow c_I=5,\qquad p=20 \Rightarrow c_I=15 (x,y)(Out,In)(x,y) \mapsto (\text{Out},\text{In})

Pooling Equilibrium

(p(5),p(15))=(15,15)(p(5),p(15))=(15,15) P(cI=5p=15)=αP(c_I=5\mid p=15)=\alpha E(π2)In=α(29)+(1α)5=534α\mathbb{E}(\pi_2)_\text{In} = \alpha(-29)+(1-\alpha)5=5-34\alpha 534α0α5345-34\alpha\geq 0 \quad\Longleftrightarrow\quad \alpha\leq \frac{5}{34} BR2(α)={Inif α534Outotherwise.BR_2(\alpha) = \begin{cases} \text{In} \quad &\text{if } \alpha\leq \frac{5}{34} \\ \text{Out} \quad &\text{otherwise.} \end{cases}

Semi-separating Equilibrium

Nash Equilibrium

Result:

The signalling game has a separating equilibrium:

{(c1=5,c1=15),(p=15,p=20)}{(p=15,p=20),Out,In}\{(c_1=5,c_1=15), (p=15,p=20)\}\mapsto\{(p=15,p=20), \text{Out},\text{In}\}

There is no pooling equilibrium and no semi-separating equilibrium.

Social Optimum

cI=5Out,cI=15Inc_I=5 \Rightarrow \text{Out},\qquad c_I=15 \Rightarrow \text{In}

Insights

Insight:

  • Price acts as a signal because diferent cost signals prefer diferent monopoly prices.
  • Separating equilibrium works when beliefs make entry unattractive against the low-cost signal and attractive against the high-cost signal.
  • Pooling fails because the high-cost incumbent has a profitable deviation.
  • No semi-separating equilibrium exists when one signal strictly prefers one signal.
← Back to Blog