Free positive multiplicative Brownian motion and the free additive convolution of semicircle and uniform distribution

Martin Auer

Free positive multiplicative Brownian motion and the free additive convolution of semicircle and uniform distribution

T0 review · 0 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A new moment formula proves the free Brownian exponential law

desk verdict Careful re-proof of a known theorem plus two genuinely new moment formulas; the delicate combinatorial identity holds up despite minor display typos. read the letter →

arxiv 2505.05984 v1 pith:CAKIFDXO submitted 2025-05-09 math.PR math.COmath.FAmath.OA

Martin Auer This is my paper · ORCID

classification math.PRmath.COmath.FAmath.OA MSC 46L5460B2060E1005A1933C15

keywords freemultiplicativeBrownianmotionadditiveconvolutionsemicircledistributionuniformmomentsStirlingnumbersofthefirstkindEgorychevmethodfractional

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper proves a closed-form formula for the moments of the free additive convolution of a semicircle and a uniform distribution: for $\mu_{\mathrm{sc},2\sqrt{t}} \boxplus \mathrm{Unif}[-t,0]$, the $n$-th moment is $m_n(t) = n! \sum_{j=\lceil n/2\rceil}^{n} \frac{t^j}{j!(1+j)!} s(1+j,n+1-j)$, where $s$ are the signed Stirling numbers of the first kind. From this formula the author gives a new proof of the known representation of the free positive multiplicative Brownian motion $\nu_t$ as the exponential image of a free additive convolution, and he derives a new fractional-moment formula for $\nu_t$. The proof works by establishing a time-inhomogeneous differential recursion for the moments and a separate combinatorial identity, proved by contour integration, that closes the recursion. A sympathetic reader should read this as a proof-by-moments of the exponential image theorem, not a numerical or heuristic confirmation.

What carries the argument

The argument is carried by two lemmas. Lemma 2.2 records that $m_n(t)$ satisfies the recursion $m_n'(t) = n\bigl(t^{-1}m_n(t) - \tfrac12\sum_{j=0}^{n-2} m_j(t)m_{n-2-j}(t)\bigr)$, obtained from the free Itô formula for a semicircular Brownian motion plus a freely independent uniform element; this turns the moment sequence into a recursively determined polynomial with known leading coefficient. Lemma 2.3 is the load-bearing combinatorial identity: for all $l,m\in\mathbb{N}$, $2^l s(1+m,1+l)$ equals a double sum over $n,k$ of products of Stirling numbers and binomial coefficients, one of them inverted. The identity is proved by Egorychev's method, a technique that evaluates combinatorial sums by writing coefficients as contour integrals, and evaluated with Kummer's confluent hypergeometric function; it supplies exactly the coefficient comparison that closes the induction proving Theorem 1.2.

What would settle it

Evaluate both sides of Lemma 2.3 for small values such as $l=2$, $m=4$ with tabulated Stirling numbers; any mismatch refutes the identity and therefore Theorem 1.2. A weaker check is to plug the Theorem 1.2 formula into Biane's moment identity (1.3) for $n=3,4,5$ and several $t$; if the two sides disagree, the claimed proof of Theorem 1.1 fails.

Watch

Extended reading notes

Core claim

The paper's central claim is Theorem 1.2: for every integer $n \ge 0$ and $t > 0$, the $n$-th moment of $\mu_{\mathrm{sc},2\sqrt{t}} \boxplus \mathrm{Unif}[-t,0]$ equals $n! \sum_{j=\lceil n/2\rceil}^{n} \frac{t^j}{j!(1+j)!} s(1+j,n+1-j)$. The author then shows that this moment formula is exactly what is needed to identify $\exp(\mu_{\mathrm{sc},2\sqrt{t}} \boxplus \mathrm{Unif}[-t/2,t/2])$ with the law $\nu_t$ of the free positive multiplicative Brownian motion: comparing the exponential generating function of these moments with Biane's known moments of $\nu_t$ forces equality of compactly supported measures, and the same calculation yields the fractional moment identity $\int_{(0,\infty)} x^\alpha\,d\nu_t = e^{\alpha t/2} {}_1F_1(1-\alpha;2;-\alpha t)$ for $\alpha \in \mathbb{C}\setminus\{0\}$.

Load-bearing premise

The load-bearing premise is Lemma 2.3: the double-sum identity with Stirling numbers and an inverted binomial coefficient is correct, and every interchange of an infinite sum with a contour integral in its proof is valid.

Editorial extensions

If this is right

Theorem 1.1 follows: $\nu_t = \exp(\mu_{\mathrm{sc},2\sqrt{t}} \boxplus \mathrm{Unif}[-t/2,t/2])$ for all $t>0$, so the free positive multiplicative Brownian motion is recovered from a free additive convolution followed by the exponential map.
Corollary 1.3 gives the fractional moments $\int_{(0,\infty)} x^\alpha\,d\nu_t = e^{\alpha t/2} {}_1F_1(1-\alpha;2;-\alpha t)$, generalizing the earlier integer-moment formulas involving Laguerre polynomials.
Remark 2.1 extends the moment formula to arbitrary $\mu_{\mathrm{sc},a} \boxplus \mathrm{Unif}[b,c]$ by scaling and a shift by a Dirac mass.
Because compactly supported measures are determined by their moments, the new calculation gives a self-contained, moment-based proof of the exponential image representation, conditional only on the combinatorial identity.
The same two-lemma structure may serve as a template for moment computations of other free additive convolutions with a semicircular component.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The coefficient identity in Lemma 2.3 looks like one member of a family: replacing the inverted binomial coefficient by another kernel in Egorychev's method could yield closed-form moments for free additive convolutions of other distributions, such as two semicircles or Marchenko–Pastur with uniform.
The fractional moment formula analytically continues in $\alpha$; one could test whether the Mellin transform $\alpha \mapsto e^{\alpha t/2}{}_1F_1(1-\alpha;2;-\alpha t)$ satisfies the free multiplicative semigroup property $\nu_s \boxtimes \nu_t = \nu_{s+t}$ directly, without passing through the integer moments.
The differential recursion in Lemma 2.2 is a closed recursion for the full polynomial $m_n(t)$, so it might give a direct route to the support or density of the convolution that bypasses the combinatorial identity altogether.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Martin Auer's paper re-proves a theorem he already proved with Voit ([1])—Theorem 1.1, the exponential-image representation of free positive multiplicative Brownian motion—but it earns its keep with two concrete new formulas. Theorem 1.2 gives a closed Stirling-number expression for moments of μ_sc,2√t ⊞ Unif[-t,0], and Corollary 1.3 turns that into a Kummer hypergeometric integral formula for fractional moments of ν_t, generalizing Biane's Laguerre expression. These are exactly the kind of formulas you want when these measures show up in computation.

The proof is essentially self-contained and honest. Lemma 2.2 imports a standard free-stochastic-calculus result; the main free work is Lemma 2.3, a double-sum Stirling identity, proved via Egorychev contours. The final moment formula matches Biane's independent moment formula (1.3), which is a solid external check. I do not see a circularity problem: the self-citation is to the original theorem, and the new derivation does not rely on it.

Soft spots, in proportion: the printed proof contains a handful of display-level inaccuracies. The beta-integral identity in Lemma 2.3's proof uses exponents n,k rather than n-1,k-1; the contour equality for j≥2 is written with a contribution that actually vanishes differently than the text claims; and the factor in (2.2) should be 2ℓ, not 2^ℓ. These are typos, not gaps—the intended identity evaluates correctly (e.g., ℓ=3,m=4 matches) and the induction in Theorem 1.2 uses the factor exactly as 2ℓ. A referee should ask for corrections before publication. The only other thing I would note is scope: this is not a breakthrough theorem; it is a careful, useful computation.

This is for free probabilists, random matrix people who need explicit moments of semicircle-uniform free convolutions, and combinatorics buffs who like Egorychev-style derivations. A serious referee should engage with it; I would send it out.

Referee Report

0 major / 4 minor

Summary. The paper gives a new, moment-based proof of the exponential representation ν_t = exp(µ_{sc,2√t} ⊞ Unif[-t/2,t/2]) for the free positive multiplicative Brownian motion. The main technical result is Theorem 1.2, a closed formula for the moments m_n(t) of µ_{sc,2√t} ⊞ Unif[-t,0] in terms of signed Stirling numbers of the first kind. The proof combines a free stochastic calculus recursion for the moments (Lemma 2.2) with a new combinatorial identity for Stirling numbers (Lemma 2.3) proved by Egorychev's contour method. From the moment formula the author derives the generating function of e^{αx} as Kummer's 1F1, which is compared with Biane's known Laguerre formula for the moments of ν_t; this yields Theorem 1.1 and the fractional-moment extension in Corollary 1.3.

Significance. The result is significant because it supplies a self-contained proof of Theorem 1.1 that avoids the random-matrix eigenvalue argument of the original proof and reduces the statement to an explicit, checkable moment identity. The moment formula in Theorem 1.2 is of independent interest for the free additive convolution of a semicircle with a uniform law, and Corollary 1.3 generalizes Biane's Laguerre moment formula to all complex orders. The proof is detailed and honest about its technical core: Lemma 2.3 is proved in full via Egorychev's method, and the final moment generating function is externally consistent with the known formula (1.3). The contour and series interchanges in the proof of Lemma 2.3 are standard Egorychev maneuvers and appear justified on the chosen contours. I find no load-bearing gap in the central claim.

minor comments (4)

[Lemma 2.3] The prefactor in (2.2) should be made unambiguous: the proof and the induction in Theorem 1.2 use the factor 2ℓ (two times ℓ), not 2^ℓ; as typeset, '2ls' invites the false reading 2^ℓ, which would make the identity incorrect.
[Proof of Lemma 2.3] The beta-integral identity displayed before (2.3) is misstated for the exponents n,k in N0: the integral ∫_0^1 t^n(1-t)^k dt is not 1/(n+k-1) times the inverse binomial coefficient (n+k choose n). The later application uses exponents n+k-1 and l+m-1-n-k, for which the evaluation 1/(l+m-1) times the inverse binomial coefficient (l+m-2 choose n+k-1) is correct; the auxiliary identity should be corrected or removed.
[Proof of Lemma 2.3] In the argument showing that the contour integrals for j≥2 vanish, the displayed equality involving (x+y)^(j-2) and x^(-j) appears to assert a false auxiliary integral; the preceding integral is nevertheless zero because log^j((1+x)/(1+y)) cancels the pole at x=y. Please correct the display or supply the intended substitution.
[Throughout] The symbol ℓ is typeset as 'l' in many displays, making it hard to distinguish from the digit 1 and from the index l; using a distinct symbol would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the new moment formula is derived independently and compared against an external benchmark.

full rationale

The derivation chain is self-contained. Theorem 1.2 is proved from a free stochastic calculus recursion (Lemma 2.2, using an external result from [10]) and a standalone Stirling-number identity (Lemma 2.3), whose Egorychev contour proof does not invoke Theorem 1.1 or Biane's moments. Theorem 1.1 is then derived by computing exponential moments from Theorem 1.2 and comparing them with Biane's known moment formula (1.3), which is an external, independent result; equality is justified by compact support and moment uniqueness. The only self-reference is [1], the original theorem being re-proved, and it is cited as the target statement rather than used as an assumption. No fitted parameter is renamed as a prediction, and no input quantity is defined in terms of the output. The minor display-level inaccuracies in the proof of Lemma 2.3 are algebraic slips, not circular steps. Therefore the paper exhibits no significant circularity.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No free parameters are fitted to data; t is the independent time variable. No new physical or mathematical entities are introduced. The proof relies on standard free probability results and one new combinatorial identity.

assumptions (5)

standard math Moment evolution ODE for free additive convolution with semicircular Brownian motion (Nikitopoulos [10], Theorem 3.5.3).
Used in Lemma 2.2 to derive the differential equation for m_n(t); a standard result in free stochastic calculus.
domain assumption Biane's moment formula (1.3): ∫ x^n dν_t(x) = e^{nt/2} (1/n) L_n^{(1)}(-nt), with L_n^{(1)} the Laguerre polynomial.
External known result from [2]; used to match moments in the proof of Theorem 1.1 and to derive Corollary 1.3.
standard math Compact support implies a probability measure is uniquely determined by its moments.
Used in the proof of Theorem 1.1 to conclude equality of measures from equality of moments; both sides have compact support.
domain assumption Existence and large-N limiting properties of free multiplicative Brownian motion (g_t) and its radial process (h_t).
Background from [8], [13], [3]; provides the definition of ν_t and its semigroup property.
standard math Generating function identities for signed Stirling numbers, the beta integral, Euler's integral transform for 1F1, and Kummer's transformation.
Used in the Egorychev-method proof of Lemma 2.3 to transform the double sum into contour integrals and hypergeometric functions.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Free positive multiplicative Brownian motion and the free additive convolution of semicircle and uniform distribution." pith.science (2026). https://pith.science/paper/CAKIFDXO

@misc{pith2026250505984,
  author       = {Pith},
  title        = {Pith review of: Free positive multiplicative Brownian motion and the free additive convolution of semicircle and uniform distribution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CAKIFDXO}},
  note         = {Machine review of arXiv:2505.05984}
}

abstract

The free positive multiplicative Brownian motion $(h_t)_{t\geq0}$ is the large $N$ limit in non-commutative distribution of matrix geometric Brownian motion. It can be constructed by setting $h_t:=g_{t/2}g_{t/2}^*$, where $(g_t)_{t\geq0}$ is a free multiplicative Brownian motion, which is the large $N$ limit in non-commutative distribution of the Brownian motion in $\operatorname{Gl}(N,\mathbb{C})$. One key property of $(h_t)_{t\geq0}$ is the fact that the corresponding spectral distributions $(\nu_t)_{t\geq0}\subset M^1((0,\infty))$ form a semigroup w.r.t. free multiplicative convolution. In recent work by M. Voit and the present author, it was shown that $\nu_t$ can be expressed by the image measure of a free additive convolution of the semicircle and the uniform distribution on an interval under the exponential map. In this paper, we provide a new proof of this result by calculating the moments of the free additive convolution of semicircle and uniform distributions on intervals. As a by-product, we also obtain new integral formulas for $\nu_t$ which generalize the corresponding known moment formulas involving Laguerre polynomials.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Hua-Pickrell diffusions and differential equations related with pseudo-Jacobi polynomials
math.PR 2026-02 conditional novelty 7.0 of 10

For Hua-Pickrell diffusions, the large-N empirical limits are independent of β, and the frozen β=∞ limits are the zeros of pseudo-Jacobi polynomials.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages · cited by 1 Pith paper

[1]

An explicit formula for free multiplicative Brownian motions via spherical functions

M. Auer and M. Voit. “An explicit formula for free multiplicative Brownian motions via spherical functions”. In: Indagationes Mathematicae (2025)

work page 2025
[2]

Free Brownian motion, free stochastic calculus, and random matrices

P. Biane. “Free Brownian motion, free stochastic calculus, and random matrices”. In: Free Probability Theory. Ed. by D.-V. Voiculescu. Vol. 12. Fields Institute Communications. Amer- ican Mathematical Society, 1997, pp. 1–19

work page 1997
[3]

Segal–Bargmann Transform, Functional Calculus on Matrix Spaces and the Theory of Semi-circular and Circular Systems

P. Biane. “Segal–Bargmann Transform, Functional Calculus on Matrix Spaces and the Theory of Semi-circular and Circular Systems”. In: Journal of Functional Analysis 144.1 (1997), pp. 232–286

work page 1997
[4]

The Brown measure of the free multiplicative Brownian motion

B. K. Driver, B. Hall, and T. Kemp. “The Brown measure of the free multiplicative Brownian motion”. In: Probability Theory and Related Fields 184.1 (Oct. 2022), pp. 209–273

work page 2022
[5]

G. P. Egorychev. Integral Representation and the Computation of Combinatorial Sums. Ed. by L. J. Leifman. Trans. by H. H. McFaden. Vol. 59. Translations of Mathematical Monographs. American Mathematical Society, 1984

work page 1984
[6]

R. L. Graham, D. E. Knuth, and O. Patashnik. Concrete Mathematics . 2nd ed. Reading, Massachusetts: Addison-Wesley, 1994, xiii+657pp

work page 1994
[7]

Log-unimodality for free positive multiplicative Brow- nian motion

T. Hasebe, Y. Ueda, and J.-C. Wang. “Log-unimodality for free positive multiplicative Brow- nian motion”. In: Colloquium Mathematicum 169 (2022), pp. 209–226. 12 REFERENCES

work page 2022
[8]

The large- N limits of Brownian motions on GLN

T. Kemp. “The large- N limits of Brownian motions on GLN”. In: International Mathematics Research Notices 2016.13 (2016), pp. 4012–4057

work page 2016

Show all 14 references

[9]

Erratum to “Itˆ o’s formula for noncommutative C 2 functions of free Itˆ o processes

E. A. Nikitopoulos. “Erratum to “Itˆ o’s formula for noncommutative C 2 functions of free Itˆ o processes””. In: Documenta Mathematica 28.5 (2023), pp. 1275–1277

2023
[10]

Itˆ o’s formula for noncommutative C 2 functions of free Itˆ o processes

E. A. Nikitopoulos. “Itˆ o’s formula for noncommutative C 2 functions of free Itˆ o processes”. In: Documenta Mathematica 27 (2022), pp. 1447–1507

2022
[11]

Egorychev Method: A Hidden Treasure

M. Riedel and H. Mahmoud. “Egorychev Method: A Hidden Treasure”. In: La Matematica 2.4 (Dec. 2023), pp. 893–933

2023
[12]

G. Szeg¨ o. Orthogonal Polynomials. 4th ed. Colloquium publications 23. American Mathemat- ical Society, 1975

1975
[13]

Limit laws for Random matrices and free products

D. Voiculescu. “Limit laws for Random matrices and free products”. In: Inventiones mathe- maticae 104.1 (Dec. 1991), pp. 201–220

1991
[14]

On the Free Convolution with a Free Multiplicative Analogue of the Normal Distribution

P. Zhong. “On the Free Convolution with a Free Multiplicative Analogue of the Normal Distribution”. In: Journal of Theoretical Probability 28.4 (Dec. 2015), pp. 1354–1379. F akult¨at Mathematik, Technische Universit ¨at Dortmund, Vogelpothsweg 87, D-44221 Dortmund, Ger- many E...

2015

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

An explicit formula for free multiplicative Brownian motions via spherical functions

M. Auer and M. Voit. “An explicit formula for free multiplicative Brownian motions via spherical functions”. In: Indagationes Mathematicae (2025)

work page 2025

[2] [2]

Free Brownian motion, free stochastic calculus, and random matrices

P. Biane. “Free Brownian motion, free stochastic calculus, and random matrices”. In: Free Probability Theory. Ed. by D.-V. Voiculescu. Vol. 12. Fields Institute Communications. Amer- ican Mathematical Society, 1997, pp. 1–19

work page 1997

[3] [3]

Segal–Bargmann Transform, Functional Calculus on Matrix Spaces and the Theory of Semi-circular and Circular Systems

P. Biane. “Segal–Bargmann Transform, Functional Calculus on Matrix Spaces and the Theory of Semi-circular and Circular Systems”. In: Journal of Functional Analysis 144.1 (1997), pp. 232–286

work page 1997

[4] [4]

The Brown measure of the free multiplicative Brownian motion

B. K. Driver, B. Hall, and T. Kemp. “The Brown measure of the free multiplicative Brownian motion”. In: Probability Theory and Related Fields 184.1 (Oct. 2022), pp. 209–273

work page 2022

[5] [5]

G. P. Egorychev. Integral Representation and the Computation of Combinatorial Sums. Ed. by L. J. Leifman. Trans. by H. H. McFaden. Vol. 59. Translations of Mathematical Monographs. American Mathematical Society, 1984

work page 1984

[6] [6]

R. L. Graham, D. E. Knuth, and O. Patashnik. Concrete Mathematics . 2nd ed. Reading, Massachusetts: Addison-Wesley, 1994, xiii+657pp

work page 1994

[7] [7]

Log-unimodality for free positive multiplicative Brow- nian motion

T. Hasebe, Y. Ueda, and J.-C. Wang. “Log-unimodality for free positive multiplicative Brow- nian motion”. In: Colloquium Mathematicum 169 (2022), pp. 209–226. 12 REFERENCES

work page 2022

[8] [8]

The large- N limits of Brownian motions on GLN

T. Kemp. “The large- N limits of Brownian motions on GLN”. In: International Mathematics Research Notices 2016.13 (2016), pp. 4012–4057

work page 2016