Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Proof of Theorem 2, p. 7627 — if ρ(E)+ρ(Eᶜ) < 1, preferences over {E,Eᶜ}-measurable acts are represented by the minimum of two linear functionals

Proved
SurelyBetter.CLB.theorem2_min_rep

by mikedeng1 · Oct 22, 2026 · Mathlib 0df444a (Lean v4.33.1)

ambiguity-aversionbiseparable-preferencescapacitychoquet-expected-utilityp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1

Let SSS be a finite set of states and let U⊆RU\subseteq\mathbb RU⊆R be a convex set containing two distinct points; UUU stands for u(X)u(X)u(X), the range of a nonconstant affine utility uuu on simple lotteries. Acts are represented by their utility profiles v:S→Uv:S\to Uv:S→U (see the definitions item for why this is exact). Let VVV be a c-linearly biseparable preference with capacity ρ\rhoρ: VVV is normalized, monotone and continuous on USU^SUS, and for every event EEE the functional W(xEy)=u(x)ρ(E)+u(y)(1−ρ(E))W(xEy)=u(x)\rho(E)+u(y)(1-\rho(E))W(xEy)=u(x)ρ(E)+u(y)(1−ρ(E)), x≿yx\succsim yx≿y, represents VVV on all binary acts. Fix an event E⊆SE\subseteq SE⊆S and write EcE^cEc for its complement.

Suppose ρ(E)+ρ(Ec)<1\rho(E)+\rho(E^c)<1ρ(E)+ρ(Ec)<1 (ambiguity aversion toward bets on EEE). Then on {E,Ec}\{E,E^c\}{E,Ec}-measurable acts, i.e. profiles hhh equal to a=u(h(E))a=u(h(E))a=u(h(E)) on EEE and b=u(h(Ec))b=u(h(E^c))b=u(h(Ec)) off EEE, the preference is represented by

m(a,b)=min⁡{ρ(E) a+(1−ρ(E)) b, (1−ρ(Ec)) a+ρ(Ec) b}:m(a,b)=\min\{\rho(E)\,a+(1-\rho(E))\,b,\ (1-\rho(E^c))\,a+\rho(E^c)\,b\}:m(a,b)=min{ρ(E)a+(1−ρ(E))b, (1−ρ(Ec))a+ρ(Ec)b}:

for all a,b,a′,b′∈Ua,b,a',b'\in Ua,b,a′,b′∈U, h≾h′h\precsim h'h≾h′ iff m(a,b)≤m(a′,b′)m(a,b)\le m(a',b')m(a,b)≤m(a′,b′).

This turns the c-linearly biseparable preference, on the two-dimensional family of bets on EEE, into a concave piecewise-linear functional; it is the first step of the proof that ambiguity aversion breaks monotonicity in optimal mixtures.

Formalization Note Acts are utility profiles S → ℝ with values in U; the hypotheses Convex ℝ U and ∃ a ∈ U, ∃ b ∈ U, a ≠ b encode that U=u(X)U=u(X)U=u(X) for a nonconstant affine uuu, and IsCLinBisep U V ρ bundles the standing MBA restriction of §4 with the representation (7) as an order equivalence over binary acts of all events. The paper justifies the representation with strict inequalities between the two linear terms; at a=ba=ba=b the two terms coincide, and the representation is stated as such.

Preamble
import Mathlib
import Definitions.Def_Supermodularity_Lattices_InducedSetOrder
import Definitions.Def_SurelyBetter_CLB_Model
Formal statement
namespace SurelyBetter.CLB
theorem theorem2_min_rep {S : Type*} [Fintype S] (U : Set ℝ) (hU : Convex ℝ U)
    (hU2 : ∃ a ∈ U, ∃ b ∈ U, a ≠ b) (V : (S → ℝ) → ℝ) (ρ : Set S → ℝ)
    (hV : IsCLinBisep U V ρ) (E : Set S) (hρ : ρ E + ρ Eᶜ < 1) :
    ∀ a ∈ U, ∀ b ∈ U, ∀ a' ∈ U, ∀ b' ∈ U,
      (V (binProf E a b) ≤ V (binProf E a' b') ↔
        min (ρ E * a + (1 - ρ E) * b) ((1 - ρ Eᶜ) * a + ρ Eᶜ * b) ≤
          min (ρ E * a' + (1 - ρ E) * b') ((1 - ρ Eᶜ) * a' + ρ Eᶜ * b')) := by sorry
end SurelyBetter.CLB
Source
Ghili & Klibanoff, If It Is Surely Better, Do It More? Implications for Preferences Under Ambiguity, Management Science 67(12) (2021), p. 7627, proof of Theorem 2 (min-representation)
Human review
  • Endorsed by Shuze Chen · Oct 23, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 23, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me