Some infinite divisibility properties of the reciprocal of planar Brownian motion exit time from a cone

With the help of the Gauss-Laplace transform for the exit time from a cone of planar Brownian motion, we obtain some infinite divisibility properties for the reciprocal of this exit time.


Introduction
Let (Z t = X t + iY t , t ≥ 0) denote a standard planar Brownian motion § , starting from x 0 + i0, x 0 > 0, where (X t , t ≥ 0) and (Y t , t ≥ 0) are two independent linear Brownian motions, starting respectively from x 0 and 0. It is well known [ItMK65] that, since x 0 = 0, (Z t , t ≥ 0) does not visit a.s. the point 0 but keeps winding around 0 infinitely often. Hence, the continuous winding process θ t = Im( t 0 dZs Zs ), t ≥ 0 is well defined. Using a scaling argument, we may assume x 0 = 1, without loss of generality, since, with obvious notation: (1) From now on, we shall take x 0 = 1. Furthermore, there is the skew product representation: where (β u + iγ u , u ≥ 0) is another planar Brownian motion starting from log 1 + i0 = 0 (for further study of the Bessel clock H, see [Yor80]). We may rewrite (2) as: One now easily obtains that the two σ-fields σ{|Z t | , t ≥ 0} and σ{β u , u ≥ 0} are identical, Bougerol's celebrated identity in law ( [Bou83,ADY97] and [Yor01] (p. 200)), which says that: We also recall several notions which will be used throughout the following text: a) A stochastic process ζ = (ζ t , t ≥ 0) is called a Lévy process if ζ 0 = 0 a.s., it has stationary and independent increments and it is almost surely right continuous with left limits. A Lévy process which is increasing is called a subordinator. b) Following e.g. [ReY99], a probability measure π on R (resp. a real-valued random variable with law π) is said to be infinitely divisible if, for any n ≥ 1, there is a probability measure π n such that π = π * n n (resp. if ζ 1 , . . . , ζ n are n i.i.d. random variables, ζ (law) = ζ 1 + . . . + ζ n ). For instance, Gaussian, Poisson and Cauchy variables are infinitely divisible. It is well-known that (e.g. [Ber96]), π is infinitely divisible if and only if, its Fourier transformπ is equal to exp(ψ), with: where b ∈ R, σ 2 ≥ 0 and ν is a Radon measure on R \ {0} such that: This expression ofπ is known as the Lévy-Khintchine formula and the measure ν as the Lévy measure. with: ] 0,1 ] | log x|µ(dx) and We remark that (10) follows immediately from (9) using the elementary Frullani formula (see e.g. [Leb72], p.6). The measure µ is called Thorin's measure associated with Γ.
We return now to the case of planar Brownian motion and the exit times from a cone. Below, we state and prove the following: Proposition 1.2 For every integer m, the function x → ϕ m (x), is the Laplace transform of an infinitely divisible random variable K; more specifically, the following decompositions hold: • for m = 2n + 1, a k e k , a k = 1 sin 2 π 2 2k−1 2n+1 ; k = 1, 2, . . . , n, • for m = 2n, where N is a centered, reduced Gaussian variable and e k , k ≤ n are n independent exponential variables, with expectation 1.
Looking at formula (6), it is also natural to consider: We note that:K admits the RHS of (6) as its Laplace transform. Hence, • for m = 2n,K with obvious notation. In Section 2 we first illustrate Proposition 1.2 for m = 1 and m = 2 ; we may also verify equation (6) by using the laws of T |θ| c , for c = π/2 and c = π/4, which are well known [ReY99]. In Section 3, we prove Proposition 1.2, where the Chebyshev polynomials play an essential role, we calculate the Lévy measure in the Lévy-Khintchine representation of ϕ m and we obtain the following asymptotic result: Proposition 1.3 With c denoting a positive constant, the distribution of T |θ| cε , for every x ≥ 0, follows the asymptotics: which, from [JRY08], is the Laplace transform of a subordinator Γ t G 1/2 , t ≥ 0 with Thorin measure that of the arc sine law, taken at t = π/2c.
Finally, we state a conjecture concerning the case where m is not necessarily an integer.

Examples
Then:φ is the Laplace transform of an exponential variable e 1 . Indeed, with t , t ≥ 0) denoting another one-dimensional BM starting from 0. Formula (6) states that: However, we know that: (20) gives:

The LHS of the previous equality
thus, we have verified directly that (20) holds. Similarly,φ is the Laplace transform of the variable N 2 2 + 2e 1 . Again, this can be shown directly; indeed, with obvious notation: Hence, formula (6) now writes, in this particular case: This is easily proven, using: T = 1 N 2 , which yields: Fubini's theorem now implies that (23) holds.
Remark 2.1 In a first draft, we continued looking at the cases: m = 3, 4, 5, 6, . . ., in a direct manner. But, these studies are now superseded by the general discussion in Section 3.

A "small" generalization
As we just wrote in Remark 2.1, before finding the proof of Proposition 1.2 (see below, Subsection 3.1), we kept developing examples for larger values of m, and in particular, we encountered quantities of the form: These quantities turn out to be the Laplace transforms of variables of the form ae + be ′ , with a, b > 0 constants and e, e ′ two independent exponential variables. In this Subsection, we characterize the polynomials P u,v (x) such that this is so.
Lemma 2.2 a) A necessary and sufficient condition for 1/P u,v to be the Laplace transform of the law of ae + be ′ , is: b) Then, we obtain: Proof of Lemma 2.2 i) 1/P u,v is the Laplace transform of ae + be ′ , then: Both u = a + b and v = ab are positive. Moreover, P u,v admits two real roots, thus ∆ ≡ u 2 − 4v ≥ 0; i.e.: (25) is satisfied.
ii) Conversely, if the two conditions (25) are satisfied, then the 2 roots of the polynomial are −1/a and −1/b. Hence, P u,v (x) = C(1 + ax)(1 + bx), where C is a constant. However, from the definition of P u,v (24), we have: P u,v (0) = 1, hence C = 1. Thus, 1/P u,v is the Laplace transform of ae + be ′ .
iii) To show b), we note that: as well as: u − √ ∆ u + √ ∆ = 4v, which finishes the proof of the second part of the Lemma.
3 A discussion of Proposition 1.2 in terms of the Chebyshev polynomials 3.1 Proof of Proposition 1.2 a) Assuming, to begin with, the validity of our Proposition 1.2, for any integer m, the function ϕ m should admit the following representation: where . In particular, P n and Q n are polynomials of degree n, each of which has its n zeros, that is (−1/a k ; k = 1, 2, . . . , n), resp. (−1/b k ; k = 1, 2, . . . , n), on the negative axis R − . It is not difficult, from the explicit expression of D m (x) = 1 2 ((G + (x)) m + (G − (x)) m ), to find the polynomials P n and Q n . They are given by the formulas: In order to prove Proposition 1.2, we shall make use of Chebyshev's polynomials of the first kind (see e.g. [Riv90] ex.1.1.1 p.5 or [KVA02] ex.25, p.195): cos (m arg cos(y)) , y ∈ [−1, 1] cosh (m arg cosh(y)) , y ≥ 1 (−1) m cosh (m arg cosh(−y)) , y ≤ 1.
(29) b) We now start the proof of Proposition 1.2 in earnest. First, we remark that: hence: with x ≥ −1, thus we are interested only in the positive zeros of T m , and we study separately the cases m odd and m even. m = 2n + 1 D 2n+1 (y) ≡ 1 + yP n (y) = T 2n+1 1 + y and the zeros of T 2n+1 are: x k = cos π 2 2k−1 2n+1 , k = 1, 2, . . . , (2n + 1). However, x k is positive if and only if k = 1, 2, . . . , n, thus: Finally: We have proved that ϕ m is infinitely divisible. In this Subsection, we shall calculate its Lévy measure. For this purpose, we shall make use of the following (recall that e k , k ≤ n are n independent exponential variables, with expectation 1): Lemma 3.1 With (c k , k = 1, 2, . . . , n) denoting a sequence of positive constants, n k=1 1 (1+c k x) is the Laplace transform of n k=1 c k e k , which is an infinitely divisible random variable with Lévy measure: dz z n k=1 e −z/c k .
We return now to the proof of Proposition 1.3 and we study separately the cases m odd and m even and we apply Lemma 3.1 with c k = a k and c k = b k respectively. m = 2n + 1 Lemma 3.1 yields that, n k=1 1 (1+a k x) is the Laplace transform of an infinitely divisible random variable with Lévy measure: Moreover: and 1 ( n k=1 (1+a k x)) 1/n , for n → ∞, converges to the Laplace transform of a variable which is a generalized Gamma convolution (GGC) with Thorin measure density: which, following the notation in [JRY08], is the Laplace transform of the variable G 1/2 which is arc sine distributed on [0, 1]. m = 2n Lemma 3.1 yields that, n k=1 1 (1+b k x) is the Laplace transform of an infinitely divisible random variable with Lévy measure: exp −z sin 2 π 2 2k − 1 2n .
Moreover 1 ( n k=1 (1+b k x)) 1/n , for n → ∞, converges to the Laplace transform of a GGC with Thorin measure density: We now express the above results in terms of the Laplace transforms ϕ m andφ m . Using the following result from [JRY08], p.390, formula (193): with 2t = m = π 2cε , with c a positive constant, together with (38) and (40), we obtain (18).
Remark: The natural question that arises now is whether the results of Proposition 1.2 could be generalized for every m > 0 (not necessarily an integer), in other words wether ϕ m (x) = 2 (G + (x)) m +(G − (x)) m is the Laplace transform of a generalized Gamma convolution (GGC, see [Bon92] or [JRY08]), that is: with Γ m (law) where f m : R + → R + and γ s is a gamma process. This conjecture will be investigated in a sequel of this article in [Vak12].