Implementation in Bayes-Nash equilibrium Tuomas Sandholm Computer Science Department

advertisement
Implementation
in Bayes-Nash equilibrium
Tuomas Sandholm
Computer Science Department
Carnegie Mellon University
Implementation in Bayes-Nash equilibrium
• Goal is to design the rules of the game (aka mechanism) so that in
Bayes-Nash equilibrium (s1, …, s|A|), the outcome of the game is
f(u1, …, u|A|)
• Weaker requirement than dominant strategy implementation
– An agent’s best response strategy may depend on others’ strategies
• Agents may benefit from counterspeculating each others’
–
–
–
–
preferences
rationality
endowments
capabilities …
– Can accomplish more than under dominant strategy implementation
• E.g., budget balance & Pareto efficiency (social welfare maximization)
under quasilinear preferences …
dAGVA expected externality mechanism
[d’Aspremont & Gerard-Varet 79; Arrow 79]
•
•
•
•
Like Groves mechanism, but sidepayment is computed based on agent’s revelation vi ,
averaging over possible true types of the others v-i
Outcome (x1, x2, ..., xk, m1, m2, ..., m|A| )
Quasilinear preferences: ui(x, m) = mi + vi(x1, x2, ..., xk)
Utilitarian setting: Social welfare maximizing choice
– Outcome s(v1, v2, ..., v|A| ) = maxx i vi(x1, x2, ..., xk)
•
Others’ expected welfare when agent i announces vi is (vi) =

v-i
p(v-i) ji vj(s(vi , v-i))
– Measures change in expected externality as agent i changes her revelation
•
•
Thrm. Assume quasilinear preferences and independently drawn valuation functions
vi. A utilitarian social choice function f: v -> (s(v), m(v)) can be implemented in BayesNash equilibrium if mi(vi)= (vi) + hi(v-i) for arbitrary function h
Unlike in dominant strategy implementation, budget balance achievable
– Intuitively, have each agent contribute an equal share of others’ payments
– Formally, set hi(v-i) = - [1 / (|A|-1)]
•
ji (vj)
Does not satisfy participation constraints (aka individual rationality constraints) in
general
– Agent might get higher expected utility by not participating
Myerson-Satterthwaite impossibility
• Avrim is selling a car to Tuomas, both are risk neutral,
quasilinear
• Each party knows his own valuation, but not the other’s
valuation
• The probability distributions are common knowledge
• Want a mechanism that is
– Ex post budget balanced
– Ex post Pareto efficient: Car changes hands iff vbuyer > vseller
– (Interim) individually rational: Both Avrim and Tuomas get
higher expected utility by participating than not
• Thrm. Such a mechanism does not exist (even if
randomized mechanisms are allowed)
– This impossibility is at the heart of more general exchange
settings (NYSE, NASDAQ, combinatorial exchanges, …) !
Proof
•
•
•
•
Seller’s valuation is sL w.p. a and sH w.p. (1-a)
Buyer’s valuation is bL w.p. b and bH w.p. (1-b). Say bH > sH > bL > sL
By revelation principle, can focus on truthful direct revelation mechanisms
p(b,s) = probability that car changes hands given revelations b and s
–
–
–
•
m(b,s) = expected price buyer pays to seller given revelations b and s
–
–
–
–
•
b E[p|b] – E[m|b]  b E[p|b’] – E[m|b’] for all b, b’
E[m|s] – s E[p|s]  E[m|s’] – s E[p|s’] for all s, s’
Suppose a=b= ½, sL=0, sH=y, bL=x, bH=x+y, where 0 < 3x < y. Now,
IR(bL): ½ x – [ ½ m(bL,sH) + ½ m(bL,sL)]  0
IR(sH): [½ m(bH,sH) + ½ m(bL,sH)] - ½ y  0
Summing gives m(bH,sH) - m(bL,sL)  y-x
Also, IC(sL): [½ m(bH,sL) + ½ m(bL,sL)]  [½ m(bH,sH) + ½ m(bL,sH)]
–
•
b E[p|b] – E[m|b]  0 for b = bL, bH
E[m|s] – s E[p|s]  0 for s = sL, sH
Bayes-Nash incentive compatibility (IC) requires
–
–
•
•
•
•
•
Since parties are risk neutral, equivalently m(b,s) = actual price buyer pays to seller
Buyer pays what seller gets paid  ex post budget balance
E[m|b] = (1-a) m(b, sH) + a m(b, sL)
E[m|s] = (1-b) m(bH, s) + b m(bL, s)
Individual rationality (IR) requires
–
–
•
Ex post efficiency requires: p(b,s) = 0 if (b = bL and s = sH), otherwise p(b,s) = 1
Thus, E[p|b=bH] = 1 and E[p|b = bL] = a
E[p|s = sH] = 1-b and E[p|s = sL] = 1
I.e., m(bH,sL) - m(bL,sH)  m(bH,sH) - m(bL,sL)
IC(bH): (x+y) - [½ m(bH,sH) + ½ m(bH,sL)]  ½ (x+y) - [½ m(bL,sH) + ½ m(bL,sL)]
–
–
I.e., x+y  m(bH,sH) - m(bL,sL) + m(bH,sL) - m(bL,sH)
So, x+y  2 [m(bH,sH) - m(bL,sL)]  2(y-x). So, 3x  y, contradiction. ■
Myerson-Satterthwaite impossibility…
• Actually, the impossibility applies to any
priors, as long as
– the priors’ supports overlap, and
– the priors don’t have gaps
• The inefficiency is caused by two-sided
private information
ADVANCED TOPICS
Full implementation: virtual implementation
• Here we don’t necessarily assume that payments are possible
• Want to implement so that the right social choice function is achieved in
all equilibria
• Virtual implementation: relax this by allowing a social choice within ε to
be implemented (for all ε > 0)
• Thm [Serrano & Vohra GEB 2005]. Consider pure strategies only, finite
type spaces, private types, and assume no-total-indifference (i.e., for each
agent and each type, there are no ties in expected utility (as long as beliefs
are updated using Bayes rule)).
A social choice function in such environments is virtually Bayesian
implementable iff it satisfies Bayesian incentive compatibility and a
condition called virtual monotonicity.
• Virtual monotonicity is weak in the sense that it is generically
satisfied in environments with at least three alternatives
• So, in most environments virtual Bayesian implementation is as
successful as it can be (i.e., incentive compatibility is the only
condition needed)
Full implementation: subgame perfection
(Hart-Moore example of Moore-Repullo mechanism)
Mechanism (assumes common knowledge between B and S)
Download