SMU H3 Notes Game TheoryGamesSMU H3

All-Pay Auction with Common Known Value

Game theory analysis: All-Pay Auction with Common Known Value.


Setup

Definition:

All-Pay Auction with Common Known Value

  • Players: nn bidders, where n2n\geq 2.
  • Strategies: Each bidder chooses a bid x0x\geq 0; In the symmetric mixed-strategy equilibrium, each bidder draws xx from the same distribution FF.

Rules

Derivation (Verification by Expected Payoff FOC)

Candidate distribution

F(x)=x1/(n1)F(x)=x^{1/(n-1)} P(win)=F(x)n1P(\text{win})=F(x)^{n-1} E(π1)=F(x)n1xE(\pi_1)=F(x)^{n-1}-x E(π1)=(x1/(n1))n1x=0E(\pi_1) =\left(x^{1/(n-1)}\right)^{n-1}-x =0

First-order condition

dE(π1)dx=0\frac{dE(\pi_1)}{dx}=0 ddx(F(x)n1x)=0\frac{d}{dx}\left(F(x)^{n-1}-x\right)=0 (n1)F(x)n2F(x)=1(n-1)F(x)^{n-2}F'(x)=1

Support

E(π1(a))=aE(\pi_1(a))=-a a=0a=0 F(x)n1x=0F(x)^{n-1}-x=0 F(x)=x1/(n1)F(x)=x^{1/(n-1)} b1/(n1)=1b^{1/(n-1)}=1 b=1b=1

Nash Equilibrium

Result:

The symmetric mixed-strategy equilibrium has support [0,1][0,1] and cumulative distribution function

F(x)=x1/(n1)F(x)=x^{1/(n-1)}

Social Optimum

01xdF(x)=101F(x)dx=101x1/(n1)dx=111/(n1)+1=1n\int_0^1 x\,dF(x) =1-\int_0^1 F(x)\,dx =1-\int_0^1 x^{1/(n-1)}\,dx =1-\frac{1}{1/(n-1)+1} =\frac{1}{n}

Diagram (Best Response Functions)

diagram

Insights

Insight:

  • All bidders pay regardless of whether they win.
  • Mixing makes every bid in [0,1][0,1] give the same expected payoff.
  • Competition dissipates the full prize value in expectation.
  • Any bid above 11 wins for sure but gives negative payoff.
← Back to Blog