Theorem 13.4: and
ProvedSupplyChainTheory.bullwhip_order_batchingTheorem 13.4. Let , and be the orders received by the supplier in period under random, positively correlated and balanced ordering respectively, in the batching model of Sect. 13.2.4 with retailers, a reorder interval of periods and i.i.d. demands, the balanced case having with . Then
- ;
- .
The orders placed to the supplier have the same mean as the demands placed to the retailers but a larger variance, so order batching produces the bullwhip effect; and correlated ordering produces the largest effect, then random, then balanced. The proof is the three variance formulas together with and .
Formalization Note The three ordering patterns are three instances of the batching model with a common , each with its own distribution of the number of ordering retailers, given as hypotheses.
import Definitions.Def_SupplyChainTheory_bullwhip
namespace SupplyChainTheory
theorem bullwhip_order_batching {Ω : Type*} [MeasurableSpace Ω] {P : MeasureTheory.Measure Ω}
[MeasureTheory.IsProbabilityMeasure P] {N R : ℕ} {mu sigma : ℝ} (hN : 0 < N) (hR : 0 < R)
(Br Bc Bb : BatchOrders P N R mu sigma)
(hXr : ∀ j : ℕ, P.real {ω | Br.X ω = j}
= (Nat.choose N j : ℝ) * (1 / R) ^ j * (1 - 1 / R) ^ (N - j))
(hXc0 : P.real {ω | Bc.X ω = 0} = 1 - 1 / R) (hXcN : P.real {ω | Bc.X ω = N} = 1 / R)
(M k : ℕ) (hk : k < R) (hNMk : N = M * R + k)
(hXbM : P.real {ω | Bb.X ω = M} = 1 - k / R)
(hXbM1 : P.real {ω | Bb.X ω = M + 1} = k / R) :
((∫ ω, Bc.supplierOrder ω ∂P) = N * mu ∧ (∫ ω, Br.supplierOrder ω ∂P) = N * mu
∧ (∫ ω, Bb.supplierOrder ω ∂P) = N * mu)
∧ (ProbabilityTheory.variance Br.supplierOrder P
≤ ProbabilityTheory.variance Bc.supplierOrder P
∧ ProbabilityTheory.variance Bb.supplierOrder P
≤ ProbabilityTheory.variance Br.supplierOrder P
∧ N * sigma ^ 2 ≤ ProbabilityTheory.variance Bb.supplierOrder P) := by sorry
end SupplyChainTheory
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.