arxiv: 2605.03703 · v2 · submitted 2026-05-05 · 🧮 math.PR · q-fin.MF

Recognition: 4 theorem links

· Lean Theorem

Scaling Limits of Bivariate Nearly-Unstable Hawkes Processes and Applications to Rough Volatility

Sohaib El Karmi

Authors on Pith no claims yet

Pith reviewed 2026-05-08 18:35 UTC · model grok-4.3

classification 🧮 math.PR q-fin.MF

keywords Hawkes processesscaling limitsrough volatilitystochastic Volterra equationsMittag-Leffler kernelsnearly unstablefunctional limit theoremtriangular cross-excitation

0 comments

The pith

Bivariate nearly-unstable Hawkes processes with distinct kernel tails converge weakly to a coupled rough stochastic Volterra system.

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

The paper proves a functional limit theorem for two Hawkes processes coupled by triangular cross-excitation when their kernels have different heavy-tail exponents. After renormalization near criticality, the intensity processes converge in law to the unique solution of a system of stochastic Volterra equations driven by two independent Brownian motions. The first component runs as an autonomous rough fractional diffusion, while the second receives input from both its own noise and the first component through a cross-kernel obtained by convolving the two associated Mittag-Leffler kernels. The result also yields an explicit short-time decorrelation rate between the components and shows that the scale-matching assumption between kernels can be relaxed by replacing the cross-kernel with a time-rescaled version.

Core claim

As the system approaches criticality, the renormalized intensity processes converge weakly to the unique solution of a coupled stochastic Volterra system driven by two independent Brownian motions. The first component evolves autonomously as a rough fractional diffusion, while the second is driven both by its own noise and by the first component through a convolution cross-kernel. This kernel, expressed as the convolution of the two associated Mittag-Leffler kernels, encodes both roughness exponents and distinguishes the limit from independent univariate limits or classical bivariate Brownian models.

What carries the argument

The convolution cross-kernel obtained from the two Mittag-Leffler kernels that encodes the distinct roughness exponents in the limiting coupled Volterra system.

If this is right

The limiting components exhibit short-time decorrelation that vanishes at a polynomial rate governed by the rougher exponent.
Dropping the scale-matching assumption replaces the limiting cross-kernel by an explicitly time-rescaled convolution kernel.
The construction supplies a tractable model for interacting rough volatility factors that carry different memory lengths.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same triangular structure and ordered tail exponents could be iterated to obtain a chain of convolution kernels for higher-dimensional systems.
The explicit decorrelation rate suggests that very short-time increments of the two components are nearly independent, which may simplify Monte Carlo schemes for the limit equations.
The convergence result supplies a natural benchmark for testing whether empirical high-frequency volatility series exhibit the predicted cross-roughness signature.

Load-bearing premise

The two kernels have distinct heavy-tail exponents and satisfy a scale-matching assumption.

What would settle it

Numerical simulation of the bivariate Hawkes processes with branching ratios approaching one, followed by checking whether the rescaled paths converge in the Skorokhod topology to the numerical solution of the corresponding Volterra system, would disprove the claim if the limiting distributions differ.

read the original abstract

We prove a functional limit theorem for a pair of nearly unstable Hawkes processes coupled through a triangular cross-excitation mechanism, when the two kernels have distinct heavy-tail exponents. This heterogeneous regime produces two different degrees of roughness and, to the best of our knowledge, had not previously been treated in the multivariate nearly unstable setting. As the system approaches criticality, the renormalized intensity processes converge weakly to the unique solution of a coupled stochastic Volterra system driven by two independent Brownian motions. The first component evolves autonomously as a rough fractional diffusion, while the second is driven both by its own noise and by the first component through a convolution cross-kernel. This kernel, expressed as the convolution of the two associated Mittag-Leffler kernels, encodes both roughness exponents and distinguishes the limit from independent univariate limits or classical bivariate Brownian models. We also derive a short-time decorrelation result showing that the functional correlation between the two limiting components vanishes at an explicit polynomial rate governed by the rougher component. Finally, we show that the scale-matching assumption is not structural: without it, the limiting cross-kernel is replaced by an explicitly time-rescaled convolution kernel. The proof combines kernel convergence, tightness, martingale identification via Rebolledo's theorem, and uniqueness for affine stochastic Volterra equations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This paper gives the first scaling limit for triangular bivariate nearly-unstable Hawkes processes with distinct heavy-tail exponents, producing an explicit convolution cross-kernel between two different roughness levels.

read the letter

The main point is that the work treats the heterogeneous bivariate case that prior multivariate nearly-unstable results had left open. As the intensities approach criticality, the renormalized pair converges to the solution of a coupled stochastic Volterra system: the first component is an autonomous rough fractional diffusion, while the second is driven by its own noise plus a convolution cross-kernel built from the two Mittag-Leffler functions. The paper also gives a short-time decorrelation rate and shows that dropping the scale-matching assumption simply inserts an explicit time-rescaling factor into the cross-kernel. These are concrete additions to the rough-volatility toolkit. The proof follows the usual route—kernel convergence to Mittag-Leffler, tightness in Skorokhod space, Rebolledo identification of the martingale part, and uniqueness for the affine Volterra equation—but applies it cleanly to the triangular structure, and the stress test finds no circularity or hidden dependence on the assumptions. The soft spots are limited. Tightness and quadratic-variation convergence are standard but still need the full manuscript to confirm every estimate; nothing in the outline suggests a load-bearing gap. The result is therefore on solid ground for what it claims. This paper is aimed at researchers who build or calibrate multivariate rough-volatility models from point-process limits. A reader already working on Hawkes scaling or fractional Volterra equations will find the explicit cross-kernel and the decorrelation statement directly usable. It deserves a serious referee because the claim is new, the method is reproducible from the given steps, and the finance application is stated clearly. I would send it out for review rather than desk-reject.

Referee Report

0 major / 3 minor

Summary. The paper proves a functional limit theorem for a bivariate system of nearly-unstable Hawkes processes coupled by a triangular cross-excitation mechanism, where the two kernels possess distinct heavy-tail exponents. As the system approaches criticality, the suitably renormalized intensity processes converge weakly in a Skorokhod space to the unique solution of a coupled affine stochastic Volterra equation driven by two independent Brownian motions; the first component is an autonomous rough fractional diffusion, while the second is driven by its own noise and by the first component through an explicit convolution cross-kernel built from the two Mittag-Leffler functions. The manuscript also establishes a short-time decorrelation result whose rate is governed by the rougher exponent and shows that the scale-matching assumption can be removed by inserting an explicit time-rescaling factor into the cross-kernel. The proof proceeds via pointwise kernel convergence, tightness, martingale identification with Rebolledo's theorem, and uniqueness for the limiting Volterra system.

Significance. If the stated convergence holds, the result supplies the first rigorous scaling limit for multivariate nearly-unstable Hawkes processes with heterogeneous roughness exponents, thereby furnishing a natural stochastic-Volterra model for rough-volatility dynamics that incorporate cross-excitation. The explicit triangular structure, the closed-form convolution cross-kernel, and the separate treatment of the non-scale-matching regime are technically useful and distinguish the limit from both independent univariate rough diffusions and classical bivariate Brownian models. The reliance on standard tools (kernel convergence to Mittag-Leffler functions, tightness, Rebolledo identification, and known uniqueness results for affine SVEs) makes the argument credible once the details are verified.

minor comments (3)

[§3] §3 (kernel convergence): the pointwise convergence of the rescaled kernels to distinct Mittag-Leffler functions is stated; please add a short appendix or reference verifying that the two heavy-tail exponents remain compatible with the integrability conditions needed for the subsequent tightness argument.
[Theorem 2.1] Theorem 2.1 (main convergence): the statement that the limit is the unique solution of the coupled Volterra system should explicitly cite the precise uniqueness theorem for affine stochastic Volterra equations that is invoked, together with a one-sentence check that the triangular cross-kernel satisfies its hypotheses.
[§4.2] §4.2 (decorrelation): the polynomial rate of short-time decorrelation is derived from the rougher exponent; confirm that the constant in the rate is uniform with respect to the initial conditions used in the tightness step.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive summary, significance assessment, and recommendation of minor revision. No major comments were raised in the report.

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper's central result is a weak-convergence theorem for rescaled bivariate Hawkes intensities to a coupled affine stochastic Volterra equation. The derivation proceeds via four independent steps: pointwise convergence of the rescaled kernels to distinct Mittag-Leffler functions, tightness of the renormalized processes in Skorokhod space, identification of the limit through Rebolledo's theorem applied to the compensated point processes (convergence of predictable quadratic variations), and uniqueness of the resulting Volterra system. These steps start from the original Hawkes intensity equations and the given kernels; the limit is obtained rather than imposed by construction, and the non-scale-matching case is treated by an explicit time-rescaling factor without altering the argument structure. No self-definitional relations, fitted inputs renamed as predictions, or load-bearing self-citations appear in the chain.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The result rests on standard probabilistic convergence theorems and uniqueness results for stochastic Volterra equations. No free parameters, ad-hoc constants, or new postulated entities are introduced in the abstract.

axioms (2)

standard math Weak convergence and tightness criteria for point processes and their compensators hold under the stated kernel conditions.
Invoked to obtain the functional limit from the renormalized Hawkes intensities.
standard math Uniqueness holds for the affine stochastic Volterra system with the given convolution kernel.
Used to identify the limit process.

pith-pipeline@v0.9.0 · 5524 in / 1415 out tokens · 69880 ms · 2026-05-08T18:35:37.445222+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

Foundation.AlphaCoordinateFixation / Cost.FunctionalEquation J_uniquely_calibrated_via_higher_derivative unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

the renormalized intensity processes converge weakly to the unique solution of a coupled stochastic Volterra system... The first component evolves autonomously as a rough fractional diffusion with Hurst index H₁ = α₁ − 1/2.
Foundation.AlphaDerivationExplicit alphaProvenanceCert (parameter-free constant derivation) unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

Mittag-Leffler kernels K_i defined through their Laplace transform: cKᵢ(z) = 1/(1 + δ̃ᵢ z^αᵢ), with αᵢ ∈ (1/2, 1) free parameters fitted to empirical equity microstructure (H ≈ 0.1).
Constants (φ-ladder) phi_golden_ratio / phi_fixed_point unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

ϱ(t) ∼ C_ϱ t^α₁ as t↓0, where α₁ is the exponent of the rougher component and C_ϱ is an explicit constant — power-law decay with non-RS exponents α₁, α₂ taken as inputs.

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

21 extracted references · 3 canonical work pages

[1]

Y. Ogata. On Lewis’ simulation method for point processes.IEEE Transactions on Information Theory, 27(1):23–31, 1981

1981
[2]

Embrechts, T

P. Embrechts, T. Liniger, and L. Lin. Multivariate Hawkes processes: an application to financial data.Journal of Applied Probability, 48(A):367–378, 2011

2011
[3]

T. W. Epps. Comovements in stock prices in the very short run.Journal of the American Statistical Association, 74(366):291–298, 1979

1979
[4]

Bayer, P

C. Bayer, P. K. Friz, and J. Gatheral. Pricing under rough volatility.Quantitative Finance, 16(6):887–904, 2016

2016
[5]

Bacry and J.-F

E. Bacry and J.-F. Muzy. Hawkes model for price and trades high-frequency dynamics.Quantitative Finance, 14(7):1147–1166, 2014

2014
[6]

Bacry, I

E. Bacry, I. Mastromatteo, and J.-F. Muzy. Hawkes processes in finance.Market Microstructure and Liquidity, 1(1):1550005, 2015

2015
[7]

El Euch and M

O. El Euch and M. Rosenbaum. The characteristic function of rough Heston models.Mathematical Finance, 29(1):3–38, 2019

2019
[8]

Gatheral, T

J. Gatheral, T. Jaisson, and M. Rosenbaum. Volatility is rough.Quantitative Finance, 18(6):933–949, 2018

2018
[9]

A. G. Hawkes. Spectra of some self-exciting and mutually exciting point processes.Biometrika, 58(1):83–90, 1971

1971
[10]

Horst and W

U. Horst and W. Xu. Functional limit theorems for nearly unstable Hawkes processes: beyond stationarity.arXiv:2401.11495, 2024

work page arXiv 2024
[11]

Jacquier and A

A. Jacquier and A. Muguruza. Correlated multi-asset rough volatility models.arXiv:1904.10488, 2019

work page arXiv 1904
[12]

Jacod and A

J. Jacod and A. N. Shiryaev.Limit Theorems for Stochastic Processes, 2nd ed. Springer, Berlin, 2003

2003
[13]

Abi Jaber, M

E. Abi Jaber, M. Larsson, and S. Pulido. Affine Volterra processes.Ann. Appl. Probab., 29(6):3155–3200, 2019

2019
[14]

Da Prato and J

G. Da Prato and J. Zabczyk.Stochastic Equations in Infinite Dimensions. Cambridge University Press, 1992

1992
[15]

Horst, W

U. Horst, W. Xu, and R. Zhang. Convergence of heavy-tailed Hawkes processes and the microstructure of rough volatility. arXiv:2312.08784, 2023

work page arXiv 2023
[16]

Jaisson and M

T. Jaisson and M. Rosenbaum. Rough fractional diffusions as scaling limits of nearly unstable heavy tailed Hawkes processes.Ann. Appl. Probab., 26(5):2860–2882, 2016. SCALING LIMITS OF BIV ARIATE NEARLY-UNSTABLE HA WKES PROCESSES AND APPLICATIONS TO ROUGH VOLATILITY 25

2016
[17]

Zolotarev,One-Dimensional Stable Distributions

V.M. Zolotarev,One-Dimensional Stable Distributions. Translations of Mathematical Monographs, Vol. 65. American Mathematical Society, Providence, RI, 1986

1986
[18]

Feller,An Introduction to Probability Theory and Its Applications, Vol

W. Feller,An Introduction to Probability Theory and Its Applications, Vol. II, 2nd ed. Wiley, New York, 1971

1971
[19]

N. H. Bingham, C. M. Goldie, and J. L. Teugels,Regular Variation. Encyclopedia of Mathematics and its Applications, Vol. 27. Cambridge University Press, Cambridge, 1987

1987
[20]

Billingsley,Convergence of Probability Measures, 2nd ed

P. Billingsley,Convergence of Probability Measures, 2nd ed. Wiley Series in Probability and Statistics. Wiley, New York, 1999

1999
[21]

Veraar, The stochastic Fubini theorem revisited.Stochastics, 84(4):543–551, 2012

M. Veraar, The stochastic Fubini theorem revisited.Stochastics, 84(4):543–551, 2012

2012