Bayesian spatial modelling framework for assessing residential flood risk in property insurance

2026-07-08

AI Summary

This study addresses the limitations of traditional insurance risk models that rely on coarse-grained geographic zoning and fail to capture fine-scale spatial heterogeneity in residential flood risk. The authors propose a policy-level, point-referenced Bayesian framework that, for the first time, applies stochastic partial differential equation (SPDE) models to flood insurance. By integrating high-resolution environmental covariates, rainfall data, and hazard maps, the approach transcends administrative boundaries. Using integrated nested Laplace approximation (INLA), the model incorporates a GLM baseline alongside several spatial specifications—including iCAR, BYM, and SPDE—to account for nonlinear covariate effects and continuous Gaussian random fields. Results demonstrate that the SPDE-based model significantly outperforms areal models in predicting claim incidence, eliminates artificial artifacts induced by arbitrary zoning, enables coherent uncertainty quantification and tail risk assessment, reveals sub-municipal risk gradients, and enhances premium allocation accuracy.

Abstract

Spatial heterogeneity in insurance risk modelling is often represented using coarse areal structures, which can obscure fine-scale patterns critical for accurate risk assessment. This study introduces a point-referenced Bayesian framework to model claim occurrence and severity at the policyholder level, avoiding reliance on predefined geographic aggregation. Drawing on a large French insurance portfolio combined with high-resolution environmental variables, rainfall records, and institutional hazard maps, we compare a benchmark GLM with several discrete Bayesian specifications, including independent random effects, intrinsic conditional autoregressive (iCAR) and Besag-York-Mollie (BYM) models, and a continuously indexed Gaussian random field constructed using the stochastic partial differential equation (SPDE) approach. Inference is performed using Integrated Nested Laplace Approximation (INLA), enabling efficient estimation of latent spatial fields and non-linear covariate effects. Our results show that accounting for spatial dependence substantially improves occurrence modelling, while gains in severity prediction are more limited. The SPDE formulation further outperforms areal models by capturing sub-municipal risk gradients and reducing artefacts induced by arbitrary geographic partitioning. By conditioning on detailed building-level attributes, we isolate the contribution of latent spatial effects, refine the interpretation of observed covariates, and improve the allocation of risk premiums across the portfolio. In addition to enhanced predictive performance, the framework provides coherent uncertainty quantification and supports tail-risk assessment. To our knowledge, this is the first application of point-referenced SPDE models to flood insurance, offering a scalable statistical alternative for pricing and managing risks with strong spatial structure.

Problem

• flood risk
• spatial heterogeneity
• property insurance
• point-referenced modeling
• risk assessment

Innovation

• point-referenced modeling
• SPDE
• Bayesian spatial modeling
• flood insurance
• INLA