Paper: arXiv 2308.01121
Authors: Cyril Bénézet, Jean-François Chassagneux, Mohan Yang
Abstract
We consider the multiple quantile hedging problem, which is a class of partial hedging problems containing as special examples the quantile hedging problem (F{“ö”}llmer & Leukert 1999) and the PnL matching problem (introduced in Bouchard & Vu 2012). In complete non-linear markets, we show that the problem can be reformulated as a kind of Monge optimal transport problem. Using this observation, we introduce a Kantorovitch version of the problem and prove that the value of both problems coincide. In the linear case, we thus obtain that the multiple quantile hedging problem can be seen as a semi-discrete optimal transport problem, for which we further introduce the dual problem. We then prove that there is no duality gap, allowing us to design a numerical method based on SGA algorithms to compute the multiple quantile hedging price.
Complexity vs Empirical Score
- Math Complexity: 9.2/10
- Empirical Rigor: 2.5/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper relies heavily on advanced mathematical concepts from optimal transport theory, stochastic control, and BSDEs, with extensive theoretical derivations and proofs, while lacking any empirical implementation details, backtests, or datasets.
Research Flowchart
flowchart TD
A["Research Goal<br>Multiple Quantile Hedging<br>in Non-Linear Markets"] --> B{"Data & Inputs"}
B --> B1["Financial Instruments<br>Option Payoffs"]
B --> B2["Market Constraints<br>Dynamic Strategies"]
B1 & B2 --> C["Key Methodology<br>Optimal Transport Framework"]
C --> C1["Reformulation<br>Monge vs Kantorovitch<br>Problem Equivalence"]
C --> C2["Duality Analysis<br>Linear Case<br>No Duality Gap"]
C1 & C2 --> D["Computational Process<br>Stochastic Gradient Ascent Algorithm"]
D --> E["Key Findings & Outcomes"]
E --> E1["Valuation<br>Hedging Price Determination"]
E --> E2["Theoretical<br>Equivalence Proof"]
E --> E3["Practical<br>Numerical Method"]
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 a new class of partial hedging problems called multiple quantile hedging and provides a numerical algorithm based on stochastic gradient ascent to compute its price.
Source sentence
... In the linear setting, we solve completely the MQH problem and design an efficient algorithm, based on Stochastic Gradient Ascent, to compute the MQH price.
Assumes 5
- There exists a uniformly bounded progressively measurable stochastic process a, b valued in R×Rm such that f(t,y,z)=a_t y + b_t z for all (t,y,z). model
Source sentence
Assumption 3.1. There exists a uniformly bounded F‑progressively measurable stochastic process
- For some ι>0, max_n |G_n| is in L2^{ι}(F_T). model
Source sentence
Assumption 3.2. For some ι ą 0, we have maxnPJNK |Gn| P L2`ιpFT , P; Rq.
- The distribution µ has full support on J^N, i.e., µ(p_n>0)=1 for all n. model
Source sentence
Assumption 3.3. The distribution µ P PpJNKq has full support on JNK, i.e. pn :“ µptnuq P p0, 1q for all n P JNK.
- The market is complete, non‑linear, allowing for some market imperfections. market
Source sentence
A general study of the MQH problem is performed in arbitrage‑free and complete but non‑linear markets, namely allowing for some market imperfections.
Show 1 more
- Under Assumptions 2.1 and 2.2, ξ↦Y_0(ξ) is convex. model
Source sentence
Proposition 2.2. Under Assumptions 2.1 and 2.2, assume further that ξ ↦ Y_0(ξ)
Data
- MQH Prices for Call Spread profiles (synthetic) synthetic unknown
Source sentence
Table 3: MQH Prices for Call Spread profiles. The targeted distribution is µ “ p1 ´ p2 ´ p3qδ1 ` p2δ2 ` p3δ3. The wealth levels are Gn “ pST ´ Kq` ´ pST ´ Knq`
Costs
Not applicable
Finds 2
- The paper proves that there is no duality gap between the Kantorovich and dual formulations of the multiple quantile hedging problem. theoretical
Source sentence
The goal of this section is to prove that there is no duality gap, i.e. that VKPpG, µq “ VDPpG, µq.
- A stochastic gradient ascent algorithm efficiently computes the multiple quantile hedging price for the tested call spread profiles. performance
Source sentence
In the linear setting, we solve completely the MQH problem and design an efficient algorithm, based on Stochastic Gradient Ascent, to compute the MQH price.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 10 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.