Paper: arXiv 2611.18677

Authors: Luca Di Persio, Francesco Mattesini

Abstract

Two credit models can quote the same bond prices today and still disagree on how information about the issuer unfolds over time. We quantify this disagreement by means of the adapted Wasserstein distance, an optimal transport distance between laws of processes consistent with the flow of information associated with them. For defaultable bond price processes, the squared distance is the smallest probability by which the two default indicators must disagree. This yields closed-form distances for rating reviews and a dynamic programming principle for multi-period rating migration. For a single review, the distance can be estimated from observed posterior survival probabilities at the parametric rate without losing calibration. For intensity models, we obtain an explicit formula for the adapted distance between bond prices that controls the error of bootstrap hedges along a quoted term structure, determine its sharp range over hazard schedules calibrated to the same prices, and show that it is stable under perturbations of the hazard rates. Finally, we complement the theoretical findings with applications to S&P transition data and Monte Carlo experiments to illustrate the size of these effects.

Complexity vs Empirical Score

  • Math Complexity: 9.0/10
  • Empirical Rigor: 7.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: This paper presents a highly mathematical and theoretically dense approach to quantifying model disagreement in credit risk using adapted Wasserstein distance. It complements this with empirical applications to S&P data and Monte Carlo experiments, demonstrating strong rigor in both theoretical development and practical illustration. The novelty lies in applying adapted optimal transport to credit risk for hedging and information flow analysis.

Research Flowchart

  flowchart TD
    A["Research Goal: Quantify disagreement between credit models on information unfolding"] --> B{"Methodology: Adapted Wasserstein Distance"};
    B --> C["Key Concepts: Optimal Transport, Default Indicators, Rating Reviews, Multi-period Rating Migration, Intensity Models"];
    C --> D{"Data/Inputs: S&P Transition Data, Observed Posterior Survival Probabilities, Hazard Schedules"};
    D --> E["Computational Processes: Closed-form distances, Dynamic Programming Principle, Parametric rate estimation, Monte Carlo experiments"];
    E --> F["Key Findings:<br>- Smallest probability of default indicator disagreement<br>- Closed-form distances for rating reviews<br>- Dynamic programming for multi-period rating migration<br>- Explicit formula for intensity models<br>- Controls bootstrap hedge error<br>- Stable under hazard rate perturbations<br>- Demonstrated with S&P data & Monte Carlo"];

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 quantifies disagreement between credit models using the adapted Wasserstein distance, derives closed‑form formulas for rating reviews and multi‑period migration, provides a dynamic programming principle, and shows how the distance can be estimated at the parametric rate and used to bound h

Source sentence
We quantify this disagreement by means of the adapted Wasserstein distance, an optimal transport distance between laws of processes consistent with the flow of information associated with them. This yields closed-form distances for rating reviews and a dynamic programming principle for multi-period rating migration.

Assumes 3

  • Time is discrete, interest rates are zero, and recovery rate is zero. market
    Source sentence
    Throughout the paper, unless specified differently, we assume that the time is discrete, the interest rates are zero, and the recovery rate is zero.
  • All probability spaces are standard Borel so that regular conditional distributions exist. model
    Source sentence
    Moreover, we assume that all probability spaces are standard Borel spaces, so that regular conditional distributions exist. 2.2.
  • The default time of the issuer is an F‑stopping time. model
    Source sentence
    Let (Ω, F, F, Q) be a filtered 10 LUCA DI PERSIO AND FRANCESCO MATTESINI probability space with filtration F = (Ft)N t=0 with F0 = {∅, Ω}, where Q is a pricing measure and Ft represents the information available to the market at date t.

Data

  • S&P transition data · S&P other unknown
    Source sentence
    Finally, we complement the theoretical findings with applications to S&P transition data and Monte Carlo experiments to illustrate the size of these effects.

Costs

Not applicable The paper reports no trading or execution result.

Finds 4

  • For defaultable bond price processes, the squared adapted Wasserstein distance equals the smallest probability by which the two default indicators must disagree. theoretical
    Source sentence
    For defaultable bond price processes, the squared distance is the smallest probability by which the two default indicators must disagree.
  • For a single rating review the distance can be estimated from observed posterior survival probabilities at the parametric rate without losing calibration. theoretical
    Source sentence
    For a single review, the distance can be estimated from observed posterior survival probabilities at the parametric rate without losing calibration.
  • In intensity models the adapted distance has an explicit formula that controls bootstrap‑hedge error, is sharp over calibrated hazard schedules, and is stable under hazard‑rate perturbations. theoretical
    Source sentence
    For intensity models, we obtain an explicit formula for the adapted distance between bond prices that controls the error of bootstrap hedges along a quoted term structure, determine its sharp range over hazard schedules calibrated to the same prices, and show that it is stable under perturbations of the hazard rates. The continuous-time analogue of (2.6) for Cox bond prices is Theorem 4.3, which identifies AW2 2
  • Empirical illustrations using S&P transition data and Monte‑Carlo experiments show that the adapted distance can be sizable and that hedging errors differ markedly between models that agree on prices. other
    Source sentence
    Finally, we complement the theoretical findings with applications to S&P transition data and Monte Carlo experiments to illustrate the size of these effects.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 17 related papers). Coverage is partial, so this is neither replication nor refutation.

Code

No repository by the authors found in the paper.

Extracted 2026-11-17 · same card as JSON · A language model chose these items: check the source sentence before relying on one.