Paper: arXiv 2509.21244
Authors: Cecilia Aubrun, Michael Benzaquen, Jean-Philippe Bouchaud
Abstract
This is the second part of our work on Multivariate Quadratic Hawkes (MQHawkes) Processes, devoted to the calibration of the model defined and studied analytically in Aubrun, C., Benzaquen, M., & Bouchaud, J. P., Quantitative Finance, 23(5), 741-758 (2023). We propose a non-parametric calibration method based on the general method of moments applied to a coarse-grained version of the MQHawkes model. This allows us to bypass challenges inherent to tick by tick data. Our main methodological innovation is a multi-step calibration procedure, first focusing on ‘‘self’’ feedback kernels, and then progressively including cross-effects. Indeed, while cross-effects are significant and interpretable, they are usually one order of magnitude smaller than self-effects, and must therefore be disentangled from noise with care. For numerical stability, we also restrict to pair interactions and only calibrate bi-variate QHawkes, neglecting higher-order interactions. Our main findings are: (a) While cross-Hawkes feedback effects have been empirically studied previously, cross-Zumbach effects are clearly identified here for the first time. The effect of recent trends of the E-Mini futures contract onto the volatility of other futures contracts is especially strong; (b) We have identified a new type of feedback that couples past realized covariance between two assets and future volatility of these two assets, with the pair E-Mini vs TBOND as a case in point; (c) A cross-leverage effect, whereby the sign of the return of one asset impacts the volatility of another asset, is also clearly identified. The cross-leverage effect between the E-Mini and the residual volatility of single stocks is notable, and surprisingly universal across the universe of stocks that we considered.
Complexity vs Empirical Score
- Math Complexity: 9.0/10
- Empirical Rigor: 7.5/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper introduces a highly advanced mathematical framework involving Multivariate Quadratic Hawkes processes, kernel calibrations, and moment method approaches, indicating high math complexity. It also demonstrates strong empirical rigor through non-parametric calibration on real high-frequency market data, identifying new cross-effects with explicit trading universe analysis.
Research Flowchart
flowchart TD
A["Research Goal<br>Calibrate MQHawkes Model<br>Non-Parametrically"] --> B["Data & Input<br>Tick-by-Tick Futures Data<br>E-Mini, TBOND, Stocks"]
B --> C["Key Methodology<br>Coarse-Grained Calibration<br>General Method of Moments"]
C --> D{"Computational Process<br>Multi-Step Calibration"}
D --> E["Step 1: Self-Effects<br>Stabilize Kernels"]
D --> F["Step 2: Cross-Effects<br>Disentangle from Noise"]
E --> G
F --> G["Pairwise Interactions<br>Bi-Variate QHawkes Only"]
G --> H["Key Findings & Outcomes<br>1. Cross-Zumbach Effects<br>2. Cross-Covariance Feedback<br>3. Cross-Leverage Effect"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a non‑parametric, multi‑step calibration method for Multivariate Quadratic Hawkes models using the general method of moments on a coarse‑grained version.
Source sentence
We propose a non‑parametric calibration method based on the general method of moments applied to a coarse‑grained version of the MQHawkes model.
Assumes 5
- All non‑equal time covariances are assumed to be zero. statistical
Source sentence
We assume here that all non‑equal time covariances are zero
- Returns are de‑drifted with zero mean (E[ri,t]=0). model
Source sentence
where all returns are de‑drifted (E[ri,t] = 0).
- Price returns are assumed to be martingales after a martingalisation step. model
Source sentence
price returns are assumed to be martingales, and many properties rely on this assumption.
- Calibration is restricted to pairwise (bi‑variate) interactions, neglecting higher‑order interactions. model
Source sentence
For numerical stability, we also restrict to pair interactions and only calibrate bi‑variate QHawkes, neglecting higher‑order interactions.
Show 1 more
- The spectral radius of the feedback matrix must be less than one for stability. statistical
Source sentence
Eq. (9) is of course only valid if the spectral radius of the feedback mat…
Data
- 1‑minute binned returns of E‑Mini and TBond futures market-prices unknown
Source sentence
Figure 1: Average intraday profiles of the squared volatility Bachelier estimates, as defined by Equation (12), for a period spanning from 2013 to 2023.
Costs
Not applicable
Finds 4
- Cross‑Zumbach effects are identified for the first time, with a strong impact of recent E‑Mini trends on other futures. theoretical
Source sentence
While cross‑Hawkes feedback effects have been empirically studied previously, cross‑Zumbach effects are clearly identified here for the first time. The effect of recent trends of the E‑Mini futures contract onto the volatility of other futures contracts is especially strong
- A new feedback coupling past realized covariance between two assets to their future volatility is discovered (E‑Mini vs TBond). theoretical
Source sentence
We have identified a new type of feedback that couples past realized covariance between two assets and future volatility of these two assets, with the pair E‑Mini vs TBOND as a case in point
- A cross‑leverage effect is clearly identified, where the sign of one asset’s return impacts another asset’s volatility, notably between E‑Mini and residual stock volatility. theoretical
Source sentence
A cross‑leverage effect, whereby the sign of the return of one asset impacts the volatility of another asset, is also clearly identified. The cross‑leverage effect between the E‑Mini and the residual volatility of single stocks is notable, and surprisingly universal across the universe of stocks that we considered.
- QGARCH‑1 factor systems operate near criticality, with values exceeding 0.955 and an average of 0.96. statistical
Source sentence
QGARCH‑1 factor systems5 are very close to 1, with all values exceeding 0.955 and an average of 0.96, indicating that the systems operate near criticality.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 24 related papers). Coverage is partial, so this is neither replication nor refutation.
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.
Cited References (63)
- Crises de liquidité endogènes dans les marchés financiers (2020)
- A nested factor model for non-linear dependencies in stock returns (2015)
- Quadratic ARCH models (1995)
- Hawkes processes and their applications to finance: a review (2018)
- Volatility is (mostly) path-dependent (2023)
- Limit theorems for nearly unstable Hawkes processes (2015)
- Quadratic Hawkes processes for financial prices (2017)
- Trades, quotes and prices: financial markets under the microscope (2018)
and 55 more
- Hawkes branching point processes without ancestors (2001)
- Point spectra of some mutually exciting point processes (1971)
- Spectra of some self-exciting and mutually exciting point processes (1971)
- Some limit theorems for Hawkes processes and application to financial statistics (2013)
- Modelling systemic price cojumps with Hawkes factor models (2015)
- Fractional Brownian motion in a nutshell (2014) — link
- Leverage effect in financial markets: The retarded volatility model (2001)
- Time reversal invariance in finance (2009)
- Generation-by-generation dissection of the response function in long memory epidemic processes (2010)
- The fine structure of volatility feedback II: Overnight and intra-day effects (2014)
- Ubiquitous power law scaling in nonlinear self-excited Hawkes processes (2021)
- The fine-structure of volatility feedback I: Multi-scale self-reflexivity (2014)
- Endogenous liquidity crises (2020)
- Some stochastic models in Eco-evolution (2016)
- Critical reflexivity in financial markets: a Hawkes process analysis (2013)
- Estimation of slowly decreasing hawkes kernels: application to high-frequency order book dynamics (2016)
- R 'e my Chicheportiche (2013)
- Self-exciting point process modeling of crime (2011)
- Epidemiological modelling of the 2005 French riots: a spreading wave and the role of contagion (2018)
- Principal regression analysis and the index leverage effect (2011)
- Prediction (2010)
- Epileptic seizures: quakes of the brain? (2010)
- Non-parametric estimation of quadratic Hawkes processes for order book events (2021)
- Hawkes processes in finance (2015)
- Hawkes processes with infinite mean intensity (2022)
- On Lewis’ simulation method for point processes (1981)
- Testing Earthquake Damage Prevention Methods with Simulated Fractal Earthquakes (1981)
- Basic models of seismicity: Temporal models (2012)
- Exogenous and endogenous price jumps belong to different dynamical classes (2022)
- New quasi-exactly solvable periodic potentials (2011)
- Multifractal random walk (2001)
- Volatility is rough (2018)
- From Rough to Multifractal volatility: the log S-fBM model (2022) — link
- Multivariate quadratic Hawkes processes—part I: theoretical analysis (2023)
- Uncovering causality from multivariate Hawkes integrated cumulants (2017)
- Branching-ratio approximation for the self-exciting Hawkes process (2014)
- The Discrete-Time 4-Factor Path-Dependent Volatility Model: Calibration under P and Q (2025)
- Modelling microstructure noise with mutually exciting point processes (2013)
- Estimators for Stationary Point Processes (1978)
- Modeling financial contagion using mutually exciting jump processes (2015)
- Multivariate Hawkes processes: an application to financial data (2011)
- On linear intensity models for mixed doubly stochastic Poisson and self-exciting point processes (1998)
- Estimation of space–time branching process models in seismology using an em–type algorithm (2008)
- A nonparametric EM algorithm for multiscale Hawkes processes (2011)
- Extending earthquakes’ reach through cascading (2008)
- Lasso and probabilistic inequalities for multivariate point processes (2015)
- Adaptive estimation for Hawkes processes; application to genome analysis (2010)
- Hawkes model for price and trades high-frequency dynamics (2014)
- Second order statistics characterization of Hawkes processes and non-parametric estimation (2014) — link
- Non-parametric kernel estimation for symmetric Hawkes processes. Application to high frequency financial data (2012)
- Statistical Learning with Sublinear Regret of Propagator Models — → 2301.05157
- Individual and collective stock dynamics: intra-day seasonalities (2011)
- Correlation structure of extreme stock returns (2001)
- Symmetry alteration of ensemble return distribution in crash and rally days of financial markets (2000)
- Unraveling Financial Market Quakes: Exploring Endogenous Volatility Dynamics in Interconnected Markets (2024)
1 of 63 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.