Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.05926

Comparing two approaches for modelling the loss given default of credit cards: Run-off triangles vs regression

Arno Botha·Henko Crewe·Marcel Muller·Janette Larney
View original source ↗Read source document ↗Export metadata ↓

Abstract

The use of run-off triangles (ROTs) is a common industry practice in estimating the loss given default (LGD) risk parameter when predicting credit losses in banking. We benchmark this industry practice using credit card data against a more sophisticated (though classical) regression-based approach, which is able to leverage various types of input variables in producing loan-level LGD-estimates. This regression-based approach can demonstrably recover the typical characteristics of the 'U-shaped' empirical LGD-distribution, which the ROT-based approach cannot do. First, we critically review the ROT-based approach and identify multiple demerits using data-driven diagnostics. We then estimate a two-stage regression-based LGD-model and favourably assess the model performance of each component (or 'stage'). Finally, we aggregate the LGD-estimates produced by each approach over time, and compare each time series to the mean empirical loss rate over time. The ROT-based aggregates diverge substantially from the empirical rate over most time periods, whilst the regression-based aggregates follow the empirical trends much closer. These results underscore the greater prediction accuracy of the regression-based LGD-model, relative to the ROT-based one. By implication, the former approach is probably better than the latter ROT-based approach when estimating the LGD under the IFRS 9 accounting framework, which prioritises accuracy.

Explore related subjects

q-fin.RMstat.AP

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Arno Botha, Henko Crewe, Marcel Muller, Janette Larney. 2026-10-05. Comparing two approaches for modelling the loss given default of credit cards: Run-off triangles vs regression. https://arxiv.org/abs/2610.05926

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Distributionally Robust Insurance under Bregman-Wasserstein Divergence

This paper investigates two optimal insurance contracting problems under distributional uncertainty from the perspective of a potential policyholder, utilizing a Bregman-Wasserstein (BW) ball to characterize the ambiguity set of loss distributions. The first problem examines an insurance demand model where the policyholder adopts an $α$-maxmin preference with Value-at-Risk (VaR). We derive the optimal indemnity function in closed form and study, both analytically and numerically, how the asymmetry inherent in BW divergence influences the optimal indemnity structure. The second problem employs a robust optimization framework, where the policyholder aims to secure robust insurance indemnity by minimizing the worst-case convex distortion risk measure while adhering to a guaranteed VaR constraint. In this context, we provide explicit characterizations of both the optimal indemnity and the worst-case distribution in closed form through a combined approach using the Lagrangian and relaxation methods. To illustrate the practical implications of our theoretical findings, we include a concrete example based on Tail Value-at-Risk (TVaR).

q-fin.RM↗
arXiv2026

A multi-view contrastive learning framework for spatial embeddings in risk modelling

Incorporating spatial information, particularly when related to climate, weather, and demographic factors, is crucial for improving underwriting precision and enhancing risk management in insurance. However, spatial data are often unstructured, high-dimensional, and difficult to integrate into predictive models. Embedding methods are needed to convert spatial data into meaningful representations for modelling tasks. We propose a novel multi-view contrastive learning framework for generating spatial embeddings that combine information from multiple spatial data sources. To train the model, we construct a spatial dataset that merges satellite imagery and OpenStreetMap features across Europe. The framework aligns these spatial views with coordinate-based encodings, producing low-dimensional embeddings that capture both spatial structure and contextual similarity. Once trained, the model generates embeddings directly from latitude-longitude pairs, enabling any dataset with coordinates to be enriched with meaningful spatial features without requiring access to the original spatial inputs. In a case study on French real estate prices, we compare models trained on raw coordinates against those using our spatial embeddings as inputs. The embeddings consistently improve predictive accuracy across generalised linear, additive, and boosting models, while providing post-hoc explainable spatial effects and demonstrating generalisation of the fitted spatial effects to regions without training observations. A second case study on flood claim counts across Belgian postal codes confirms that the embeddings improve territorial risk classification in an insurance context.

q-fin.RM↗
arXiv2026

VaR at Its Extremes: Impossibilities and Conditions for One-Sided Random Variables

Value-at-Risk (VaR) may reward diversification at some probability levels and penalize it at others. We study the two extremal regimes in which one effect prevails at every level simultaneously, for sums of one-sided risks. For risks supported on $[0,\infty)$, VaR sub-additivity at all levels, exact VaR additivity and co-monotonicity are equivalent without any integrability assumption; for such risks this removes the finite-mean hypothesis of Imamura and Kato (2026), which cannot be dropped for two-sided risks. Global super-additivity, in turn, is equivalent to first-order stochastic dominance of the aggregate over its co-monotonic counterpart. We factorize this condition through the maximum of independent copies of the margins into negative simplex dependence (NSD), a condition on the joint law that is strictly weaker than negative lower orthant dependence, and simplex dominance (SD) of the aggregator $Φ(x_1,\dots,x_n)=\sum_{i=1}^n x_i\log F_{X_i}(x_i)$, which for continuous margins holds whenever each reciprocal risk $1/X_i$ has a decreasing failure rate on average. The criterion admits heterogeneous margins, yields a lower bound on the diversification penalty that depends on the margins only, is exact for unit Fréchet margins, and accommodates positively and tail-dependent risks such as common-shock models. Independently of any factorization, a vector that is VaR super-additive but not VaR additive has at least two infinite-mean components, and this bound is sharp. A reflection principle extends all results to finite endpoints: lower-bounded risks are never strictly sub-additive, upper-bounded risks are never strictly super-additive, and compactly supported risks are neither.

q-fin.RM↗