Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Theorem 6, p. 893 — for stationary (Y_n) in a Banach algebra with E(log‖Y₁‖)⁺ < ∞, n⁻¹ log‖Y₁⋯Y_n‖ → ξ a.s. and E(ξ) = lim n⁻¹E log‖Y₁⋯Y_n‖

Proved
KingmanSubadditive.BanachAlgebra.theorem_6

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

banach-algebralyapunov-exponentp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1random-productssubadditive-ergodic-theorem

Let B\mathfrak BB be a real or complex Banach algebra and let (Yn)n≥1(Y_n)_{n\ge1}(Yn​)n≥1​ be a stationary sequence of random elements of B\mathfrak BB on a probability space (Ω,F,P)(\Omega,\mathcal F,P)(Ω,F,P). Suppose that

E{(log⁡∥Y1∥)+}<∞.(2.3.3)E\{(\log\|Y_1\|)^+\}<\infty. \tag{2.3.3}E{(log∥Y1​∥)+}<∞.(2.3.3)

Then the limit

ξ=lim⁡n→∞n−1log⁡∥Y1Y2⋯Yn∥(2.3.4)\xi=\lim_{n\to\infty}n^{-1}\log\|Y_1Y_2\cdots Y_n\| \tag{2.3.4}ξ=n→∞lim​n−1log∥Y1​Y2​⋯Yn​∥(2.3.4)

exists in −∞≤ξ<∞-\infty\le\xi<\infty−∞≤ξ<∞ with probability one, and

E(ξ)=lim⁡n→∞n−1E{log⁡∥Y1Y2⋯Yn∥}.(2.3.5)E(\xi)=\lim_{n\to\infty}n^{-1}E\{\log\|Y_1Y_2\cdots Y_n\|\}. \tag{2.3.5}E(ξ)=n→∞lim​n−1E{log∥Y1​Y2​⋯Yn​∥}.(2.3.5)

The quantity eξe^{\xi}eξ is a stochastic analogue of the spectral radius: if every YnY_nYn​ equals a fixed element yyy, then eξe^\xieξ is the spectral radius of yyy by Gelfand's formula. For k×kk\times kk×k matrices with any matrix norm the theorem is the Furstenberg–Kesten theorem on products of random matrices.

Formalization Note log⁡∥⋅∥\log\|\cdot\|log∥⋅∥ is extended-real valued (log⁡∥0∥=−∞\log\|0\|=-\inftylog∥0∥=−∞) and all expectations are extended expectations ∫f+ dP−∫f− dP∈[−∞,∞)\int f^+\,dP-\int f^-\,dP\in[-\infty,\infty)∫f+dP−∫f−dP∈[−∞,∞). The conclusion asserts: ξ\xiξ is a measurable [−∞,∞][-\infty,\infty][−∞,∞]-valued random variable; almost surely ξ<∞\xi<\inftyξ<∞ and n−1log⁡∥Y1⋯Yn∥→ξn^{-1}\log\|Y_1\cdots Y_n\|\to\xin−1log∥Y1​⋯Yn​∥→ξ in [−∞,∞][-\infty,\infty][−∞,∞]; E(ξ+)<∞E(\xi^+)<\inftyE(ξ+)<∞; and n−1E{log⁡∥Y1⋯Yn∥}→E(ξ)n^{-1}E\{\log\|Y_1\cdots Y_n\|\}\to E(\xi)n−1E{log∥Y1​⋯Yn​∥}→E(ξ) in [−∞,∞][-\infty,\infty][−∞,∞], which asserts both that the limit exists and that it equals E(ξ)E(\xi)E(ξ). Stationarity is equality of joint laws on BN\mathfrak B^{\mathbb N}BN with the product σ\sigmaσ-algebra; each YnY_nYn​ is strongly measurable and B\mathfrak BB need not be separable. B\mathfrak BB is a complete unital normed ring and a normed algebra over R\mathbb RR or C\mathbb CC.

Preamble
import Mathlib
import Definitions.Def_KingmanSubadditive_BanachAlgebra_Process
import Definitions.Def_KingmanSubadditive_BanachAlgebra_RandomProduct
Formal statement
namespace KingmanSubadditive.BanachAlgebra

open MeasureTheory Filter Topology

/-- **Theorem 6, §2.3, p. 893** (Kingman, *Subadditive ergodic theory*, Ann. Probab.
1(6):883–899 (1973), DOI 10.1214/aop/1176996798). "Let `(Y_n)` be a stationary sequence taking
values in a (real or complex) Banach algebra `𝔅`, and suppose that (2.3.3)
`E{(log ‖Y₁‖)⁺} < ∞`. Then (2.3.4) `ξ = lim_{n→∞} n⁻¹ log ‖Y₁ Y₂ ⋯ Y_n‖` exists in
`−∞ ≤ ξ < ∞` with probability one, and (2.3.5) `E(ξ) = lim_{n→∞} n⁻¹ E{log ‖Y₁ Y₂ ⋯ Y_n‖}`."

**Formalization Note.** `log ‖·‖` is the extended-real logarithm (`log ‖0‖ = −∞`), so
`log ‖Y₁ ⋯ Y_n‖`, `ξ` and all means are in `EReal`; expectations are the extended expectation
`eMean` (`∫ f⁺ − ∫ f⁻`, possibly `−∞`). The conclusion asserts: `ξ` is a measurable
`EReal`-valued random variable; almost surely `ξ < +∞` and `n⁻¹ log ‖Y₁ ⋯ Y_n‖ → ξ` in `[−∞, ∞]`;
`E(ξ⁺) < ∞` (so `E(ξ)` is defined); and the sequence `n⁻¹ E{log ‖Y₁ ⋯ Y_n‖}` converges to `E(ξ)`
in `[−∞, ∞]` (both existence of the limit and its value). The term at `n = 0` does not affect the
limits. The product is ordered `Y₁ Y₂ ⋯ Y_n`. Each `Y_n` (`n ≥ 1`) is strongly measurable and the
sequence is stationary in the joint-law sense (`IsStationarySeq`); `𝔅` need not be separable. -/
theorem theorem_6 {𝕜 𝔅 Ω : Type*} [RCLike 𝕜] [NormedRing 𝔅]
    [NormedAlgebra 𝕜 𝔅] [CompleteSpace 𝔅] [MeasurableSpace 𝔅] [BorelSpace 𝔅]
    [MeasurableSpace Ω] (P : Measure Ω) [IsProbabilityMeasure P] (Y : ℕ → Ω → 𝔅)
    (hY : ∀ n : ℕ, 1 ≤ n → StronglyMeasurable (Y n)) (hstat : IsStationarySeq P Y)
    (hlog : ∫⁻ ω, (logNorm (Y 1 ω)).toENNReal ∂P < ⊤) :
    ∃ ξ : Ω → EReal, Measurable ξ ∧
      (∀ᵐ ω ∂P, ξ ω ≠ ⊤ ∧
        Tendsto (fun n : ℕ => (((n : ℝ)⁻¹ : ℝ) : EReal) * logNorm (prodRange Y 0 n ω))
          atTop (𝓝 (ξ ω))) ∧
      ∫⁻ ω, (ξ ω).toENNReal ∂P < ⊤ ∧
      Tendsto (fun n : ℕ => (((n : ℝ)⁻¹ : ℝ) : EReal) *
          eMean P (fun ω => logNorm (prodRange Y 0 n ω)))
        atTop (𝓝 (eMean P ξ)) := by sorry

end KingmanSubadditive.BanachAlgebra
Source
Kingman, Subadditive ergodic theory, Ann. Probab. 1(6):883–899 (1973), DOI 10.1214/aop/1176996798, p. 893, §2.3, Theorem 6, (2.3.3)–(2.3.5)
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 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