Lemma 5, p. 20 — the largest optimal grid bid b⋆(v) is non-decreasing in the value v
ProvedFirstPriceLearning.Stochastic.lemma_5first-price-auctionsmonotone-comparative-staticsp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1
Let be any CDF on , the expected reward (2), and the bid grid (6) with . For let be the largest maximiser of over . Then
With contexts ordered by value, this is the second condition of Definition 3 for the quantized first-price auction: the optimal bid moves up with the value.
Preamble
import Mathlib import Definitions.Def_FirstPriceLearning_Stochastic_Auction
Formal statement
namespace FirstPriceLearning.Stochastic
open MeasureTheory
/-- Lemma 5 (p. 20): on the bid grid `𝒜 = {1/K, …, 1}` of (6), for any HOB law `G` on `[0, 1]`,
the largest maximiser `b⋆(v)` of `b ↦ R(v, b) = (v - b) G(b)` over `𝒜` is non-decreasing in
`v ∈ [0, 1]`. -/
theorem lemma_5 (K : ℕ) (hK : 1 ≤ K) (μG : Measure ℝ) [IsProbabilityMeasure μG]
(hG : μG (Set.Icc (0 : ℝ) 1)ᶜ = 0) :
MonotoneOn (largestMaximizer μG (gridA K)) (Set.Icc 0 1) := by sorry
end FirstPriceLearning.Stochastic
Source
Han, Zhou, Weissman, Optimal No-regret Learning in Repeated First-price Auctions, arXiv:2003.09795v7, p. 20, Lemma 5
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.