Section A.5 —
ProvedGradedReserve.candidate_lossauctionsbayes-nash-equilibriummechanism-designp2o-gran-per-chapterp2o-plan-paperp2o-v1revenue-guarantee
Let ( real), and . The anchored loss of the candidate is
With (25) this gives the optimal anchored revenue of Theorem 3.1.
Differences from the source:
- encoding: F(v) = v^k with k ≥ 1 → powerDensity k (density k v^{k−1} on [0,1]); k real, Real.rpow; reason: the same law for every real k ≥ 1
Preamble
import Mathlib import Definitions.Def_GradedReserve_Cfun import Definitions.Def_GradedReserve_cZero import Definitions.Def_GradedReserve_pMin import Definitions.Def_GradedReserve_powN import Definitions.Def_GradedReserve_vbar import Definitions.Def_GradedReserve_xStar open MeasureTheory ProbabilityTheory Set Filter Topology
Formal statement
namespace GradedReserve
/-- **The candidate's anchored loss** (Chen–Wang, *Optimal Auctions for Anchored Bidders under a Rational Revenue Guarantee*, manuscript (STOC 2027 submission), Section A.5, p.19, PDF 20): direct integration gives
`C[x*] = (N/(N+1)) (c₀ v̄)^{N+1} p/(p+N+1)` for `p > p_min`. -/
theorem candidate_loss (k : ℝ) (hk : 1 ≤ k) (n : ℕ) (hn : 2 ≤ n) (p : ℝ) (hp : pMin k n < p) :
Cfun k n (xStar k n p) =
powN k n / (powN k n + 1) * (cZero k n p * vbar k n p) ^ (powN k n + 1)
* (p / (p + powN k n + 1)) := by sorry
end GradedReserve
Source
Chen–Wang, Optimal Auctions for Anchored Bidders under a Rational Revenue Guarantee, manuscript (STOC 2027 submission; PDF sha256 6b1da440…1bab), Section A.5, p.19 (PDF 20)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.