Paper: arXiv 2310.13797
Authors: Manuel Hasenbichler, Benjamin Joseph, Gregoire Loeper, Jan Obloj, Gudmund Pammer
Abstract
We develop a numerical method for the martingale analogue of the Benamou-Brenier optimal transport problem, which seeks a martingale interpolating two prescribed marginals which is closest to the Brownian motion. Recent contributions have established existence and uniqueness for the optimal martingale under finite second moment assumptions on the marginals, but numerical methods exist only in the one-dimensional setting. We introduce an iterative scheme, a martingale analogue of the celebrated Sinkhorn algorithm, and prove its convergence in arbitrary dimension under minimal assumptions. In particular, we show that convergence holds when the marginals have finite moments of order \(p > 1\), thereby extending the known theory beyond the finite-second-moment regime. The proof relies on a strict descent property for the dual value of the martingale Benamou–Brenier problem. While the descent property admits a direct verification in the case of compactly supported marginals, obtaining uniform control on the iterates without assuming compact support is substantially more delicate and constitutes the main technical challenge.
Complexity vs Empirical Score
- Math Complexity: 9.0/10
- Empirical Rigor: 2.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper introduces a novel, convergent iterative algorithm for a martingale optimal transport problem, requiring advanced stochastic analysis, convex duality, and proofs of convergence in arbitrary dimensions, placing high demand on mathematical complexity. However, the work is purely theoretical, presenting a convergence proof under minimal assumptions without any numerical experiments, backtests, or empirical data analysis, resulting in low empirical rigor.
Research Flowchart
flowchart TD A["Research Goal:<br>Develop a numerical method<br>for the Martingale Benamou-Brenier problem"] --> B["Key Methodology:<br>Introduce a martingale analogue<br>of the Sinkhorn algorithm"] B --> C["Data/Inputs:<br>Prescribed marginals<br>Finite moment assumptions p > 1"] C --> D["Computational Process:<br>Iterative dual descent<br>proving strict descent property"] D --> E["Key Findings:<br>Convergence in arbitrary dimension<br>Extension beyond finite-second-moment regime"]
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 introduces an iterative Martingale Sinkhorn algorithm, a martingale analogue of the Sinkhorn algorithm, to solve the martingale Benamou–Brenier optimal transport problem in arbitrary dimension.
Source sentence
We introduce an iterative scheme, a martingale analogue of the celebrated Sinkhorn algorithm, and prove that it yields a Bass potential in arbitrary dimension under minimal assumptions.
Assumes 2
- The marginals are assumed to have finite moments of order p > 1. statistical
Source sentence
In particular, we show that this holds when the marginals have finite moments of order $p > 1$, thereby extending the known theory beyond the finite-second-moment regime.
- A strict descent property is directly verified when the marginals are compactly supported. model
Source sentence
While the descent property admits a direct verification in the case of compactly supported marginals, obtaining uniform control on the iterates without assuming compact support is substantially more delicate and constitutes the main technical challenge.
Data
- Empirical measures constructed from independent samples of each marginal synthetic public
Source sentence
For all three examples we construct empirical measures from 20 000 independent samples from each marginal.
Costs
Not applicable
Finds 5
- The Martingale Sinkhorn algorithm yields a Bass potential in arbitrary dimension under minimal assumptions. theoretical
Source sentence
We introduce an iterative scheme, a martingale analogue of the celebrated Sinkhorn algorithm, and prove that it yields a Bass potential in arbitrary dimension under minimal assumptions.
- The dual value of the martingale Benamou–Brenier problem exhibits a strict descent property along the algorithm. theoretical
Source sentence
The proof relies on a strict descent property for the dual value of the martingale Benamou--Brenier problem.
- Uniform control of the iterates without compact support is achieved via a delicate technical argument. theoretical
Source sentence
While the descent property admits a direct verification in the case of compactly supported marginals, obtaining uniform control on the iterates without assuming compact support is substantially more delicate and constitutes the main technical challenge.
- Each iterate preserves the quadratic moment condition, ensuring all iterates lie in P₂(ℝᵈ). theoretical
Source sentence
Since hθ2 ≃1 2| · |2 at infinity, we conclude by the transport relation Z |x|2 αθ2(dx) = Z |∇hθ2|2 dµ0 < ∞, so that each iterate αθ2 lies in P2(Rd) and the quadratic moment condition is preserved along the scheme.
- In a synthetic example, the Bass measure may have substantially heavier tails than its marginals, lacking a finite second moment despite compact support of µ₀ and µ₁. theoretical
Source sentence
In particular, its Bass measure α does not admit a finite second moment, although µ0 and µ1 are compactly supported.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 21 related papers). Coverage is partial, so this is neither replication nor refutation.
6 papers in the archive : Robust and Fast Bass local volatility
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.