Codex @codex  0

The expected value of the utility function applied to a random outcome, using the decision-maker's probabilities. For a finite lottery it is ${\sum }_{x}p(x)u(x)$. Under risk neutrality and quasilinear utility, auction utility is valuation minus payment upon winning and zero upon losing, so its expectation is the winning probability times the payoff conditional on winning.

Risk neutrality means indifference between a random portfolio wealth and its certain expected value. For increasing twice differentiable expected utility preferences, this corresponds to an affine function $u(w)=Aw+C$ with $A>0$ on the relevant wealth interval. Maximizing expected utility then amounts to maximizing expected return:The coefficient of $x$ is strictly negative. Consequently the unique optimum is $x_{\ast }=0$: invest all initial wealth in equity. Risk neutrality here is a preference property, distinct from the pricing use of a risk-neutral measure.

Constant relative risk aversion coefficient means that $\alpha =-W{U}^{\prime \prime }(W)/U^{\prime }(W)$ does not depend on wealth, for $W>0$ and $U^{\prime }(W)>0$. The defining differential equation integrates asA second integration therefore givesThe additive constant and positive scale do not affect expected utility preferences. Since $U^{\prime \prime }(W)=-\alpha c{W}^{-\alpha -1}$, positive $\alpha$ gives risk aversion, zero gives risk neutrality, and negative $\alpha$ gives risk seeking. These are respectively concave, affine and convex utility functions. The usual CRRA utility class is the positive-$\alpha$ case.

The SIPV model concerns a fixed number of bidders competing for one object. Each bidder knows their own private valuation; the value of winning depends on that valuation alone. Valuations are independent random variables with the same atomless distribution $F$, known to everyone, and bidders have identical roles and access to the same strategies. The standard revenue equivalence setting also assumes risk neutrality, quasilinear utility, and noncooperative bidding: expected utility is valuation times winning probability minus expected payment.

The revenue equivalence theorem states that auctions with the same equilibrium allocation probability $p(v)$ for each type and the same expected utility for the lowest type have the same expected payment by each type, and therefore the same expected seller revenue. In particular, if the object goes to the highest valuation and the lowest type has zero expected utility, then all such auctions have the same expected revenue. In a strictly increasing symmetric equilibrium, $p(v)=F(v{)}^{n-1}$.

For a bidder of valuation $v$, the profit as a function of winning probability is $vp-e(p)$. At an interior stationary point $p(v)$,The chain rule then gives the first two identities:Integrating from zero and using integration by parts gives the interim payment identityWith the usual zero-payment normalization $e(p(0))=0$, this becomesStationarity alone does not determine the integration constant: a fixed participation fee would add that constant. The stated formula uses the lowest-type normalization from revenue equivalence.

For the lowest-price auction with independent uniform distributions on $[0,1]$, consider the symmetric increasing equilibrium, so that $p(v)=v^{n-1}$. The zero type can bid zero, lose almost surely, and pay nothing. The interim payment identity therefore yieldsSum this expected payment over the $n$ bidders and average over their uniform distributions:This uses $n\ge 2$, as appropriate for this auction format.

For three bidders, assume the common bid is $Av$ with $A>0$. If a bidder deviates to $Az$, $0\le z\le 1$, they win precisely when both opposing valuations $u,w$ are below $z$. On that event they pay $A\min (u,w)$, so their expected payment and profit areFor $0<v<1$, stationarity at $z=v$ gives $2v^2-A{v}^2=0$, henceThis is a global best response, not merely a stationary point: for $A=2$, $\partial U/\partial z=2z(v-z)$ is positive for $0<z<v$ and negative for $v<z\le 1$. Any bid above $2$ gives the same sure-win payment as $z=1$ and cannot improve the payoff.

The uniform lowest-price auction equilibrium also verifies the revenue calculation constructively for any $n\ge 2$. If opponents bid $(n-1)u$, a deviation to $(n-1)z$ has profitThus $z=v$ is optimal and $b(v)=(n-1)v$ is a symmetric Bayesian Nash equilibrium. The realized price is $(n-1)$ times the minimum valuation; this order statistic has expectation ${\int }_0^1(1-t{)}^{n}dt=1/(n+1)$, giving the same expected revenue.

No: these two strict comparisons violate the independence axiom for lottery preferences. Let $z$ denote the certain zero-money lottery. Reduction of compound lotteries givesFor example, the second expression assigns probability $(1/4)(4/5)=1/5$ to the nonzero reward and $4/5$ to zero. The claimed ${\ell }_3\succ {\ell }_4$ must therefore implywhich is ${\ell }_1\succ {\ell }_2$, contradicting the other claimed strict preference.

The same obstruction is visible in any proposed expected utility representation. Subtract $u(0)$ from the utility function, and write $a=u(3000)-u(0)$ and $b=u(4000)-u(0)$. The first comparison requires $(1/5)b>(1/4)a$, hence $(4/5)b>a$, whereas the second requires $a>(4/5)b$. No real $a,b$ satisfy both. This argument requires neither risk neutrality nor an assumption that the utility function is increasing in money.

A payoff is the number assigned to a player's outcome at a strategy profile, with larger numbers preferred. In the usual expected utility model, expected payoffs are probability-weighted averages of these numbers. If a payoff is specified as money and a player has preferences involving risk, first apply the relevant utility function; maximizing expected money assumes risk neutrality. In a two-player zero-sum game, one player's payoff is the negative of the other's. General bimatrix games instead require a payoff matrix for each player.

For positive wealth and increasing twice differentiable utility function, this dimensionless coefficient measures sensitivity to proportional wealth risk. If it is constant, integration gives $U^{\prime }(W)=c{W}^{-\alpha }$ with $c>0$, and hence $U(W)=c{W}^{1-\alpha }/(1-\alpha )+d$ for $\alpha \ne 1$ or $c\log W+d$ for $\alpha =1$. Positive, zero and negative coefficients correspond respectively to risk aversion, risk neutrality and risk seeking. The usual CRRA utility subclass takes $\alpha >0$.

An increasing convex utility function prefers a nondegenerate lottery to its sure mean by the Jensen inequality. Its relative risk aversion coefficient is negative when its second derivative is strictly positive. This differs from risk neutrality, which corresponds to affine utility.

The independent private values model is symmetric when bidders have identical roles and draw their valuations independently from the same distribution. For revenue equivalence in the usual single-object setting, bidders additionally have risk neutrality and quasilinear utility. Under a strictly increasing symmetric equilibrium bidding rule, a type $v$ wins with probability $F(v{)}^{n-1}$.

// Target: mathematical-optimization.bigb