Theorem 7.1.5 — Brownian paths are Hölder continuous of every exponent below 1/2
ProvedDurrettProbability.brownian_holderLet be a Brownian motion on a probability space : its finite-dimensional laws are those of Brownian motion and almost every path is continuous. Let be a non-negative real with .
Then for -almost every and every there is a constant with
The null set is chosen before : a single path satisfies the estimate on every bounded initial interval, with a constant that may grow with .
Formalization Note Hölder continuity is stated on rather than on because the global statement is false — the constant genuinely must depend on the interval. The estimate is Mathlib's Hölder-on-a-set predicate, phrased with extended distances, which on the reals is the displayed inequality.
The exponent is a non-negative real and both bounds are strict, so the borderline exponent is excluded, as it must be: Durrett records that for each fixed . No claim is made about exponents above , where the conclusion is false.
import Mathlib import Definitions.Def_DurrettProbability_Brownian open Filter MeasureTheory ProbabilityTheory open scoped NNReal Topology
namespace DurrettProbability
theorem brownian_holder {Ω : Type*} [MeasurableSpace Ω] {P : Measure Ω}
[IsProbabilityMeasure P] {B : ℝ≥0 → Ω → ℝ} (hB : IsBrownianReal B P)
{γ : ℝ≥0} (hγ : 0 < γ) (hγ2 : γ < 1 / 2) :
∀ᵐ ω ∂P, ∀ T : ℝ≥0, ∃ C : ℝ≥0,
HolderOnWith C γ (fun t : ℝ≥0 => B t ω) (Set.Iic T) := by sorry
end DurrettProbabilityConfirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.