for
ProvedMartingale.norm_exp_I_mul_sub_one_add_I_mul_leA quantitative second-order estimate: for real with ,
It is the complex exponential estimate , valid for , specialised to the purely imaginary , where .
Where it is used. McLeish's proof of the martingale central limit theorem replaces the characteristic function — which cannot be factored, since the summands are not independent — by the comparison product , whose expectation is exactly by the martingale property. This lemma is what quantifies the substitution: each factor agrees with to first order, with an error of size .
That the error is quadratic is precisely what the argument needs. The first-order terms cancel exactly, so the ratio carries no linear term and its logarithm begins at ; summing over produces , which the limiting-variance hypothesis converts into the Gaussian exponent . Any comparison factor agreeing with only to zeroth order would leave a linear term and destroy the identification.
The hypothesis is the range in which the underlying exponential estimate holds with constant ; in the application it is supplied by the negligibility of the increments, which makes uniformly small for large .
import Mathlib.Analysis.SpecialFunctions.Exponential import Mathlib.Analysis.SpecialFunctions.Complex.Circle
theorem Martingale.norm_exp_I_mul_sub_one_add_I_mul_le (x : ℝ) (hx : |x| ≤ 1) :
‖Complex.exp (Complex.I * x) - (1 + Complex.I * x)‖ ≤ x ^ 2 := by sorry