Theorem 11.22 -- equilibrium_price_exists_iff_feasible
OpenDiscreteConvex.EconomicEquilibriumB.equilibrium_price_exists_iff_feasiblediscrete-convex-analysisdiscrete-geometry
Theorem 11.22 (p.344). There exists an equilibrium price vector (for the allocation ) iff the linear system (11.43)/(11.44) is feasible.
Scope note. The book's own trailing algorithmic remarks (that the smallest/largest equilibrium price vectors can be found by solving a shortest-path problem, and the resulting polynomial-time algorithm for checking equilibrium existence) are computational/complexity content and are omitted; the mathematical "iff feasibility" content is placed in full. See HARD.md.
(Murota, Discrete Convex Analysis, SIAM 2003, DOI 10.1137/1.9780898718508, p.344, Theorem 11.22.)
Preamble
import Mathlib import Definitions.Def_DiscreteConvex_EconomicEquilibriumB_EquilibriumPriceSet import Definitions.Def_DiscreteConvex_EconomicEquilibriumB_EquilibriumPricePolyhedronE import Definitions.Def_DiscreteConvex_EconomicEquilibriumB_MNaturalConcave import Definitions.Def_DiscreteConvex_EconomicEquilibriumB_MNaturalConvexC import Definitions.Def_DiscreteConvex_EconomicEquilibriumB_IsOptimalAllocation
Formal statement
namespace DiscreteConvex.EconomicEquilibriumB
open Classical
open scoped Pointwise
variable {K : Type*} [Fintype K] [DecidableEq K]
/-- Theorem 11.22 (p.344). There exists an equilibrium price vector (for the allocation `(x,y)`)
iff the linear system (11.44)/(11.43) is feasible. Murota, *Discrete Convex Analysis*, SIAM 2003, §11.5 assumes the utilities
M♮-concave, the cost functions M♮-convex and `(x,y)` an optimal allocation of the associated
MSFP2, and its bounds are read in `EReal`: with one good, one consumer with `U(0) = U(1) = 0`,
`U(2) = 5` and `-∞` elsewhere and one producer with `C(0) = C(2) = 0`, `C(1) = 1`, the equilibrium
price set is empty while the bounds taken with `unbotD 0`/`untopD 0` give `{0}`. -/
theorem equilibrium_price_exists_iff_feasible {H L : Type*} [Fintype H] [Fintype L] [Nonempty H]
[Nonempty L] (U : H → (K → ℤ) → WithBot ℝ) (C : L → (K → ℤ) → WithTop ℝ) (x : H → (K → ℤ))
(y : L → (K → ℤ)) (hU : ∀ h, MNaturalConcave (U h)) (hC : ∀ l, MNaturalConvexC (C l))
(hopt : IsOptimalAllocation U C x y) :
(EquilibriumPriceSet U C x y).Nonempty ↔ (EquilibriumPricePolyhedronE U C x y).Nonempty := by sorry
end DiscreteConvex.EconomicEquilibriumB
Source
Murota, Discrete Convex Analysis, SIAM 2003, DOI 10.1137/1.9780898718508, p.344, Theorem 11.22
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.