Dual evaluation of performance and fairness from machine learning models for non-life insurance pricing

Simple item page
dc.contributor.authorIsrani, Tarun
dc.contributor.authorWolsztynski, Eric
dc.contributor.authorDaly, Linda
dc.contributor.authorCondon, John
dc.date.accessioned2026-02-11T14:50:04Z
dc.date.available2026-02-11T14:50:04Z
dc.date.issued2026-01-29
dc.description.abstractAn increasing number of reports highlight the potential of machine learning (ML) methodologies over the conventional generalised linear model (GLM) for non-life insurance pricing. In parallel, national and international regulatory institutions are accentuating their focus on pricing fairness to quantify and mitigate algorithmic differences and discrimination. However, comprehensive studies that assess both pricing accuracy and fairness remain scarce. We propose a benchmark of the GLM against mainstream regularised linear models and tree-based ensemble models under two popular distribution modelling strategies (Poisson-gamma and Tweedie), with respect to key criteria including estimation bias, deviance, risk differentiation, competitiveness, loss ratios, discrimination and fairness. Pricing performance and fairness were assessed simultaneously on the same samples of premium estimates for GLM and ML models. The models were compared on two open-access motor insurance datasets, each with a different type of cover (fully comprehensive and third-party liability). While no single ML model outperformed across both pricing and discrimination metrics, the GLM significantly underperformed for most. The results indicate that ML may be considered a realistic and reasonable alternative to current practices. We advocate that benchmarking exercises for risk prediction models should be carried out to assess both pricing accuracy and fairness for any given portfolio.en
dc.description.sponsorshipThis publication has emanated from research conducted with the financial support of Research Ireland under Grant number 12/RC/2289-P2.
dc.description.versionPublished Version
dc.format.extent36
dc.format.mimetypeapplication/pdfen
dc.identifier.articleide5
dc.identifier.authororcidIsrani, Tarun
dc.identifier.authororcidWolsztynski, Eric§0000-0001-9770-9769
dc.identifier.authororcidDaly, Linda§0009-0002-2982-6407
dc.identifier.authororcidCondon, John
dc.identifier.citationIsrani, T, Wolsztynski, E, Daly, L & Condon, J 2026, 'Dual evaluation of performance and fairness from machine learning models for non-life insurance pricing', British Actuarial Journal, vol. 31, e5, pp. 1-36. https://doi.org/10.1017/S1357321725100317
dc.identifier.doi10.1017/S1357321725100317
dc.identifier.endpage36
dc.identifier.issn1357-3217
dc.identifier.journaltitleBritish Actuarial Journal
dc.identifier.otherORCID: /0009-0002-2982-6407/work/205466759
dc.identifier.otherORCID: /0000-0001-9770-9769/work/205467181
dc.identifier.startpage1
dc.identifier.urihttps://hdl.handle.net/10468/18515
dc.identifier.volume31
dc.language.isoen
dc.publisherCambridge University Press
dc.relation.urihttps://www.cambridge.org/core/journals/british-actuarial-journal/article/dual-evaluation-of-performance-and-fairness-from-machine-learning-models-for-nonlife-insurance-pricing/0F19EB9EAB28A6581368EEFA3D1F69BD
dc.rights© 2026, the Authors. Published by Cambridge University Press on behalf of The Institute and Faculty of Actuaries. This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
dc.rights.accessrightsopen access
dc.rights.licensenameAttribution 4.0 International
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.statusPeer reviewed
dc.subjectActuarial pricing
dc.subjectFairness
dc.subjectGeneral insurance
dc.subjectMachine learning
dc.subjectNon-life insurance
dc.subjectPricing bias
dc.subjectPricing structure
dc.subjectProtected variables
dc.subjectRate making
dc.subject[Maths]
dc.titleDual evaluation of performance and fairness from machine learning models for non-life insurance pricingen
dc.typeArticle (peer-reviewed)
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
dual-evaluation-of-performance-and-fairness-from-machine-learning-models-for-non-life-insurance-pricing.pdf
Size:
4.09 MB
Format:
Adobe Portable Document Format
Download
Collections
School of Mathematical Sciences