Proposition 4(d) — under production diseconomy the retailer earns more under sequential, the manufacturers under concurrent information contracting
ProvedInfoSharing.Diseconomy.sequential_vs_concurrentLet be a signal model with parameters , and let , , . Then the pricing game has an equilibrium family, and for every pricing-equilibrium family:
- concurrent contracting has an outcome, and sequential contracting has a subgame-perfect equilibrium for either first manufacturer ;
- for every concurrent outcome , every and every SPE with first mover ,
where is the retailer's profit after the side payments and the manufacturers' total profit net of them; 3. if moreover , both inequalities are strict.
The retailer prefers the protocol in which she can exploit the manufacturers' competition for information; the supply chain's manufacturers prefer the one that treats them alike.
Formalization Note. The pricing stage is the Bayesian game of §4.1 on the signal model; the ex ante profits , are those of an arbitrary pricing-equilibrium family, not the §4.2 closed forms. Wholesale strategies are measurable, square-integrable functions of the signal value (constants for an uninformed manufacturer); the production cost is with and ; manufacturers are indexed by . The paper says "higher". Below neither protocol sells information and all profits coincide, and at an SPE may still sell nothing, so "higher" is stated as everywhere and strictly above .
import Mathlib import Definitions.Def_InfoSharing_Shared_IsSignalModel import Definitions.Def_InfoSharing_Diseconomy_IsConcurrentOutcome import Definitions.Def_InfoSharing_Shared_IsSequentialSPE open MeasureTheory open InfoSharing.Shared
namespace InfoSharing.Diseconomy
/-- Proposition 4(d), p. 254 (production diseconomy): the retailer's profit is higher under
sequential information contracting, whereas the total profit of the manufacturers is higher
under concurrent information contracting.
The pricing stage has an equilibrium family; for every such family both contracting games have
outcomes; every concurrent outcome and every SPE of the sequential game (either first mover)
compare weakly as stated, and strictly when `c > (√2 − 1)/(1 + φ)` (below it neither protocol
sells information and all profits coincide). -/
theorem sequential_vs_concurrent {Ω : Type*} [MeasurableSpace Ω] (μ : Measure Ω)
(θ Y : Ω → ℝ) (a b c φ σ β : ℝ) (hmodel : IsSignalModel μ θ Y σ β) (hφ : 0 < φ)
(hb : 0 < b) (hc : 0 < c) :
(∃ E : (Fin 2 → Status) → PricingProfile, IsPricingEqFamily μ θ Y a b c φ β E) ∧
∀ E : (Fin 2 → Status) → PricingProfile, IsPricingEqFamily μ θ Y a b c φ β E →
let P := payoffTable μ θ Y a b c φ E
(∃ (T : ℝ) (X : Fin 2 → Status), IsConcurrentOutcome P T X) ∧
(∀ k : Fin 2, ∃ s : SeqStrategy, IsSequentialSPE P k s) ∧
∀ (T : ℝ) (X : Fin 2 → Status) (k : Fin 2) (s : SeqStrategy),
IsConcurrentOutcome P T X → IsSequentialSPE P k s →
concRetailer P T X ≤ seqRetailer P k s ∧
seqManufacturersTotal P k s ≤ concManufacturersTotal P T X ∧
((Real.sqrt 2 - 1) / (1 + φ) < c →
concRetailer P T X < seqRetailer P k s ∧
seqManufacturersTotal P k s < concManufacturersTotal P T X) := by sorry
end InfoSharing.Diseconomy
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.