Showing 529 of 529 articles
No articles match that filter.
April 2026
174 articles
-
05 Apr 2026Applying fairness constraints, calibration corrections, and drift monitoring as sequential post-hoc steps is how most UK pricing teams work. It is also archi...
-
05 Apr 2026SS1/23 is in active enforcement for banks. Its monitoring principles — operating boundaries, outcome analysis, documented escalation — are the benchmark the ...
-
05 Apr 2026UK motor NCR is at 111% (EY Q4 2024). The market is at or near technical floor. Here is how to identify underpriced segments using loss cost trending, A/E mo...
-
05 Apr 2026A companion post to our March coverage of arXiv:2603.17106. This one walks through the code: where exactly in a UK proxy-based fairness audit does the bias f...
-
05 Apr 2026SS1/23 applies to banks, not insurers — but its monitoring principles are the benchmark auditors use regardless. This post maps the five SS1/23 validation pr...
-
05 Apr 2026FCA cohort 2 is live as of April 2026. This post maps the four things a submission actually needs to include — outcome monitoring, proxy bias detection, unce...
-
05 Apr 2026The FCA is consulting in Q2 2026 on exactly this question. Every layer in a DA chain — insurer, MGA, broker, PCW — is expected to evidence outcomes, not just...
-
05 Apr 2026FCA AI Live Testing cohort 2 opened in January 2026. Testing is live from April. Firms in scope for AI underwriting and pricing need to demonstrate four thin...
-
05 Apr 2026insurance-conformal v1.3.1 adds ConditionalCoverageAssessor — a tool for detecting and decomposing conditional coverage failures in conformal prediction inte...
-
04 Apr 2026A complete Python tutorial for Whittaker-Henderson smoothing of insurance rating tables. Replace your Excel moving average or SAS graduation with automatic R...
-
04 Apr 2026You don't need a JBA licence to build a materially better flood model. A French study on 968,000 policies shows which open data sources actually move the nee...
-
04 Apr 2026A practical Python tutorial for telematics pricing: load raw GPS trip data, classify driving regimes with a Hidden Markov Model, and produce GLM-ready risk f...
-
04 Apr 2026A hands-on Python tutorial for insurance pricing analysts on survival analysis and lapse modelling. Covers Kaplan-Meier, Weibull AFT, mixture cure models, cu...
-
04 Apr 2026Regression by composition — the framework that generalises GAMLSS and transformation models — suffers from a subtle non-identifiability when you stack multip...
-
04 Apr 2026Chouldechova (2017) proved that when group base rates differ, no classifier can simultaneously achieve calibration within groups, equal false positive rates,...
-
04 Apr 2026Conformal Risk Control Assumes Your Loss Decreases With Interval Width. Most Insurance Losses Don't.Conformal risk control (Angelopoulos et al. ICLR 2024) requires monotone loss functions for its finite-sample guarantees. The Winkler score, two-sided regula...
-
04 Apr 2026The FCA confirmed its motor finance redress scheme on 30 March 2026 — £7.5bn, 12.1m agreements, going back to 2007. The enforcement pattern is not motor-fina...
-
04 Apr 2026Richman and Wüthrich's March 2026 paper (arXiv:2603.11660) proves that aggregate chain-ladder produces materially biased ultimate estimates on liability line...
-
04 Apr 2026C. Evans Hedges (Lemonade, December 2025) derives the first closed-form formula connecting model discrimination to expected loss ratio. LRE translates a corr...
-
04 Apr 2026How to set up insurance model monitoring in Python from scratch: PSI, Gini drift, and A/E with the insurance-monitoring library. Know when to redeploy, recal...
-
04 Apr 2026EquiPy (Fernandes Machado, Charpentier et al.) does distributional fairness correction via Wasserstein barycenters. insurance-fairness does proxy discriminat...
-
04 Apr 2026Step-by-step Python tutorial for running an insurance fairness audit. Covers proxy discrimination detection, exposure-weighted bias metrics, FCA Consumer Dut...
-
04 Apr 2026glum fits the Tweedie GLM in seconds. Here is how our libraries handle everything around it: distribution-free prediction intervals, PSI/CSI drift monitoring...
-
04 Apr 2026End-to-end GLM frequency model in Python using freMTPL2 from OpenML. Data prep, exposure handling, glum fitting, deviance residuals, actual vs expected, and ...
-
04 Apr 2026A hands-on tutorial on GAM insurance pricing in Python using the insurance-gam library. Covers EBM tariff construction, shape function extraction, GLM compar...
-
04 Apr 2026A single freMTPL2 motor pipeline running through insurance-gam, insurance-conformal, insurance-monitoring, insurance-fairness, and insurance-governance. No o...
-
04 Apr 2026Le, Denis and Hebiri (arXiv:2604.02017, April 2026) show that enforcing demographic parity over the full prediction distribution is both accuracy-costly and ...
-
04 Apr 2026A practical Python tutorial on credibility theory for insurance pricing analysts. Covers Buhlmann-Straub model, the insurance-credibility library, UK motor e...
-
04 Apr 2026A step-by-step tutorial on conformal prediction for insurance Python models, specifically the frequency-severity decomposition. Covers the calibration subtle...
-
04 Apr 2026Seo and Lim (arXiv:2604.01629, April 2026) show that standard empirical Bayes methods inflate FDR to 0.25–0.35 when the prior is misspecified — even when the...
-
04 Apr 2026A hands-on tutorial on causal inference for insurance pricing in Python using the insurance-causal library. Covers double machine learning (DML), CatBoost nu...
-
04 Apr 2026Nayak's Calibrated Credit Intelligence (arXiv:2603.06733) is a credit paper, but UK insurers should read it. It addresses uncertainty quantification, fairnes...
-
04 Apr 2026Boulevard regularisation turns EBM shape functions into kernel ridge regression estimates, giving valid CLT-based confidence intervals. We show why this matt...
-
04 Apr 2026You deploy a conformal pricing model and check its coverage every month. After twelve months, you see a coverage failure. Is the model broken, or did you jus...
-
04 Apr 2026Habermann et al. (ICLR 2025, arXiv:2408.13230) train a neural network to approximate posteriors for hierarchical Bayesian models — once. After training, any ...
-
04 Apr 2026How to run actuarial model validation in Python for UK insurance pricing models. Covers Solvency II Article 120 and Consumer Duty requirements, the five-test...
-
03 Apr 2026Standard quantile regression overfits badly on small insurance segments. A new closed-form result from Zhang, Mao and Wang (2026) gives a distributionally ro...
-
03 Apr 2026A February 2026 paper provides the first statistically valid confidence intervals for global SHAP feature importance. We explain what changes for UK insuranc...
-
03 Apr 2026The Lindholm-Richman-Tsanakas-Wüthrich EJOR 2026 paper gives pricing teams something they did not previously have: an instance-level proxy discrimination sco...
-
03 Apr 2026Frequency model monitoring has a growing toolkit. Reserve models — which target specific quantiles of the loss distribution, not the mean — have almost nothi...
-
03 Apr 2026A stable Gini coefficient is not evidence that a model is performing well. It is evidence that the model is still ranking risks in the same order. Score deco...
-
03 Apr 2026Zanzouri et al. (NAAJ 2025) benchmark four ML severity models inside the QPP framework. The tau adjustment is elegant. CatBoost was missing from the comparis...
-
03 Apr 2026INLA — the method that made Bayesian spatial GLMMs tractable in R — now has a proper Python package. 92–278x faster than PyMC, no R dependency. We explain wh...
-
03 Apr 2026Liu & Meng's PowerBurr adds a fourth parameter to Burr XII that decouples body shape from tail heaviness. We explain the coupling trap, what the fix actually...
-
03 Apr 2026Shen et al. (arXiv:2603.24875) gives valid CIs for Lasso-selected Poisson GLMs. That fixes your frequency model. UK motor pricing is Poisson × Gamma — and th...
-
03 Apr 2026insurance-gam v0.3.0 ships PostSelectionGLM and DataSplitPostSelectionGLM: two classes that produce valid confidence intervals for Poisson frequency models a...
-
03 Apr 2026The Pool Adjacent Violators Algorithm solves an O(N) monotonicity problem with no parametric assumptions. It appears in three distinct insurance pricing cont...
-
03 Apr 2026Boucher & Coulibaly (arXiv:2502.11788) prove that offset and ratio exposure handling are equivalent for Poisson frequency models — but diverge for Tweedie pu...
-
03 Apr 2026NeuralGaussianMixture is now in insurance-distributional v0.4.0. The question is not whether it can fit bimodal severity — it can. The question is whether yo...
-
03 Apr 2026insurance-distributional now has five distributional model classes. NeuralGaussianMixture is the newest and the most demanding. A routing guide: which model ...
-
03 Apr 2026Negative log-likelihood is a proper scoring rule. So why does NLL training collapse a Mixture Density Network to a single component? The answer is in the los...
-
03 Apr 2026The hunger-for-bonus effect biases your NCD frequency relativities. It also biases your severity model. The two errors partially offset each other — but the ...
-
03 Apr 2026The Bonferroni correction for joint frequency-severity prediction sets is conservative by construction. Braun et al. (arXiv:2507.20941) show that covariance ...
-
03 Apr 2026The per-transition Lindholm correction for multi-state models is sound. But arXiv:2602.04791 has four gaps that matter before you build on it: no accuracy de...
-
03 Apr 2026A new 173-page handbook from MARSAIL documents one Thai insurer's computer vision system for vehicle damage assessment. It contains one genuinely useful data...
-
03 Apr 2026UK personal lines generates hundreds of millions of competitor quotes per year. The industry treats them as competitive positioning data. They are, in fact, ...
-
03 Apr 2026Thibodeau et al. build a multi-firm market simulator and train an RL social planner to design fairness tax schedules. The collusion result stops the paper co...
-
03 Apr 2026A working Python implementation of Goffard, Piette & Peters (2025) ABC-SMC market-based ratemaking. Forty lines of vectorised numpy, PAVA via scipy, loss rat...
-
03 Apr 2026An FCA Research Note (December 2025) found a £28 unexplained ethnicity residual across six million motor policies. Your pricing team cannot measure it becaus...
-
03 Apr 2026Miao & Pesenti's KL discrimination-insensitive result is theoretically clean. Deploying it in a production GLM-based pricing system is not. The paper is sile...
-
03 Apr 2026Most fairness sections in model validation reports say nothing. The KL discrimination-insensitive result from Miao & Pesenti (2026) gives you a precise, defe...
-
03 Apr 2026Miao & Pesenti (2026) derive the nearest fair probability measure in KL divergence. Their existence theorem is useful for FCA Consumer Duty attestation. Our ...
-
03 Apr 2026Holtan (2001) showed that the NCD reporting threshold falls when interest rates rise — the NPV of future premium penalties shrinks, so policyholders become m...
-
03 Apr 2026A new empirical paper from Richman and Wüthrich shows that individual claims models cut Mack RMSEP by 44% on accident data and correct severe bias on liabili...
-
03 Apr 2026Lee et al. (arXiv:2602.02398) prove that standard hurdle-Poisson models with bivariate normal random effects can violate credibility order — your frequency e...
-
03 Apr 2026Separate peril GLMs routinely disagree with each other by 3–8% at cover level. insurance-reconcile applies premium-weighted MinTrace to make your hierarchy c...
-
03 Apr 2026GAMformer produces GAM shape functions in a single forward pass, no hyperparameter search, sub-second inference. For insurance pricing today: three hard bloc...
-
03 Apr 2026Insurance fraud ML papers routinely overstate their results through five avoidable errors: wrong evaluation metric, no external baseline, random train/test s...
-
03 Apr 2026Most pricing teams have model documentation. Almost none have per-decision audit trails. The FCA's Q2 2026 AI evaluation will test exactly this gap. Here is ...
-
03 Apr 2026Everyone is buying CyberCube or Kovrr. The open data suggests entity-level cyber frequency scoring is feasible — but the honest numbers (r = 0.36, US-only, n...
-
03 Apr 2026PSI > 0.2 and A/E > 1.15 are industry folklore, not statistics. Conformal SPC replaces them with calibrated p-values that have a finite-sample false alarm ra...
-
03 Apr 2026IJCNN 2025 paper arXiv:2509.00846 introduces Causal SHAP: it uses causal discovery to estimate a DAG, then computes SHAP values that respect causal structure...
-
03 Apr 2026The hardest part of fitting a GPD is picking the threshold. A new Bayesian nonparametric approach eliminates the choice entirely — and tells you what fractio...
-
03 Apr 2026Nieto-Barajas (arXiv:2602.07228) proposes a Bayesian nonparametric mixture of Shifted Gamma-Gamma distributions that eliminates EVT threshold selection entir...
-
03 Apr 2026Orihara, Momozaki & Sugasawa (arXiv:2506.04868) produce a Bayesian posterior over the ATE by tilting the product of independent posteriors to satisfy the DR ...
-
02 Apr 2026UK motor bodily injury severity is structurally bimodal. A GammaGBM fits one mode between two humps, understating the 95th percentile by 30-40%. NeuralGaussi...
-
02 Apr 2026If you have called RetentionUpliftModel with outcome='survival', your model silently ran as binary. There is no pip-installable Python package for survival C...
-
02 Apr 2026Wang & Yu (arXiv:2603.01232) characterise which risk measures are submodular — mathematically encoding that diversification benefits are real but bounded. ES...
-
02 Apr 2026Das (arXiv:2603.24640) establishes stochastic ordering results for minimum and maximum claims across heterogeneous portfolios with random claim counts. Usefu...
-
02 Apr 2026Bimonte et al. show that age-specific Shapley weights across 15 mortality models outperform any single model at 10-20 year horizons — exactly the range that ...
-
02 Apr 2026SHAPInference (in development for shap-relativities): asymptotically valid confidence intervals on global SHAP feature importance using de-biased U-statistic...
-
02 Apr 2026Most pricing teams eyeball calibration plots or track PSI. Neither tells you whether observed miscalibration is statistically significant. ScoreDecomposition...
-
02 Apr 2026insurance-monitoring v1.2.0 adds ScoreDecompositionTest: HAC-robust p-values on the miscalibration and discrimination components of any scoring rule. When yo...
-
02 Apr 2026When a challenger GBM outperforms the production GLM on deviance, the improvement is often entirely in miscalibration — meaning the GLM could be recalibrated...
-
02 Apr 2026insurance-optimise v0.5.0 adds RobustReinsuranceOptimiser — closed-form and numerical optimal proportional reinsurance cession under model uncertainty, based...
-
02 Apr 2026insurance-glm-tools v0.2.0 ships RobustMMDGLM — a Gamma GLM that automatically downweights large losses and selects features via L1, based on Kang & Kang (20...
-
02 Apr 2026Avanzi, Richman, Wüthrich et al. (arXiv:2601.07637) treat individual claim development as a Markov decision process, using Soft Actor-Critic to revise outsta...
-
02 Apr 2026Boonen & Ghossoub (arXiv:2602.14223) prove that when a reinsurer sets terms first — Bowley/Stackelberg — the treaty is mathematically guaranteed to be worse ...
-
02 Apr 2026Kirke (arXiv:2603.15664) applies Quantum Amplitude Estimation to catastrophe insurance tail-risk pricing and claims quadratic speedup over classical Monte Ca...
-
02 Apr 2026insurance-fairness v1.2.0 adds PrivatizedFairPricer: discrimination-free pricing when the sensitive attribute is privatised via local differential privacy. B...
-
02 Apr 2026Burr XII's body and tail are controlled by the same parameters — you can't fix one without breaking the other. Liu & Meng's PowerBurr adds a fourth parameter...
-
02 Apr 2026PopulationSamplingReserve lands in insurance-severity v0.4.0. IBNR as a missing-data problem: the AIPW doubly-robust estimator hedges your bets between the c...
-
02 Apr 2026A new paper models hazard functions as solutions to nonlinear ODEs, producing shapes no standard parametric family can match. The maths is genuinely interest...
-
02 Apr 2026Yang et al. (arXiv:2603.27672) fix mode collapse in Mixture Density Networks by adding an analytic Energy Score term to the training objective. The contribut...
-
02 Apr 2026insurance-monitoring v1.0.0 adds ModelMonitor with check_gmcb and check_lmcb — separate tests for global and local calibration drift, wired into a three-way ...
-
02 Apr 2026Samuel Clark's MDMx (arXiv:2603.20518) applies Tucker tensor decomposition to HMD mortality data across 50 countries, producing coherent forecasts, disruptio...
-
02 Apr 2026insurance-fairness v0.8.0 adds LocalizedParityCorrector and LocalizedParityAudit — enforcement of demographic parity at pricing tier boundaries, not just on ...
-
02 Apr 2026Beauchemin & Khoury (arXiv:2603.07825) benchmark 51 LLMs on Quebec insurance regulatory certification questions. Passing insurance exams is the wrong success...
-
02 Apr 2026insurance-optimise v0.6.0 adds LinearRiskSharingPool: Cramér-Lundberg ruin analysis for community-based risk pools, based on Denuit, Flores-Contró & Robert (...
-
02 Apr 2026insurance-survival v0.5.0 adds LifetimeBoundsCalculator — closed-form worst- and best-case mortality bounds between integer ages for variable annuity and gua...
-
02 Apr 2026insurance-conformal v1.2.0 adds LCPModelSelector: locally adaptive conformal model and score selection that gives per-prediction tighter intervals while main...
-
02 Apr 2026When you acquire a portfolio or enter a scheme, your pricing model was fitted on a different risk population. Weill and Wang (2026) give a kernel GLM framewo...
-
02 Apr 2026A new mortality model from Liu & Zhou (2026) shows that cause-specific shocks decay heterogeneously — some fast, some slow. The analogy to UK claims inflatio...
-
02 Apr 2026Richman & Wüthrich's one-shot individual claims reserving framework (arXiv:2603.11660) shows that a simple OLS model — using case estimates as the primary fe...
-
02 Apr 2026Shankar & Cohen automate GAM structure search using NSGA-II evolutionary algorithms. The idea is legitimate; the problem is that EBM already does this better...
-
02 Apr 2026Asadi & Li's Generative Adversarial Regression (arXiv:2603.08553) frames scenario generation as a minimax problem between a generator and an adversarial poli...
-
02 Apr 2026Flood Re ends in 2039. From that date, 350,000+ currently subsidised properties need risk-reflective pricing. We work through Moriah et al. 2026's sequential...
-
02 Apr 2026The FCA's pure protection market study interim report landed in January 2026. The final report is due Q3 2026. For income protection pricing teams, the centr...
-
02 Apr 2026The FCA published its first consolidated insurance priorities report on 24 February 2026. It replaces individual portfolio letters and signals where regulato...
-
02 Apr 2026Pricing teams treat fairness as a single slider between accuracy and parity. NSGA-II reveals it is a landscape with multiple competing criteria. Here is what...
-
02 Apr 2026Kong, Liu & Yang prove that standard conformal coverage guarantees degrade unevenly when protected attributes are absent at test time. With post-ECJ gender p...
-
02 Apr 2026insurance-governance v0.3.0 adds ExplainabilityAuditTrail: a per-prediction audit log that records SHAP values, fairness flags, and plain-language summaries ...
-
02 Apr 2026ExpectedShortfallRegressor in insurance-quantile v0.5.0 implements direct ES regression via the i-Rock method. No quantile detour. Sandwich SEs. Directly rel...
-
02 Apr 2026Gevorgyan et al. propose exchangeable multi-task Gaussian processes for causal effect estimation in staggered-adoption designs. The method handles nonlinear ...
-
02 Apr 2026Liu & Luger (arXiv:2603.02357) build a semiparametric VaR/ES forecaster that models scale dynamics through quantile differences rather than GARCH variance eq...
-
02 Apr 2026A new paper (arXiv:2504.09396) uses PPO reinforcement learning with a CVaR constraint to manage a reserve buffer. The framing is interesting — but this is no...
-
02 Apr 2026insurance-optimise v0.7.0 adds ConvexRiskReinsuranceOptimiser — analytically optimal multi-line reinsurance contracts under CVaR or variance constraints, bas...
-
02 Apr 2026Coverage-matched conformal rules can have wildly different deployment profiles. Three operational metrics that matter for pricing deployment.
-
02 Apr 2026Marginal 90% coverage can hide severe undercoverage for specific risk profiles. ConditionalCoverageAssessor — new in insurance-conformal v1.2.0 — quantifies ...
-
02 Apr 2026Most pricing teams treat acquisition and retention as separate problems. CLV-aware ratemaking integrates expected policy duration, cross-sell probability, an...
-
02 Apr 2026RMSE on closed claims is biased. Corollary 3 of Taggart, Loveday & Louis (arXiv:2603.14835) proves mean settlement time is not scoreable under right-censorin...
-
02 Apr 2026Nieto-Barajas (arXiv:2602.07228, 2026) proposes a Bayesian nonparametric mixture of Scaled Generalised Gaussian distributions that eliminates threshold selec...
-
02 Apr 2026Two January 2026 arXiv papers formalise what motor actuaries have always known informally: NCD creates rational incentives to suppress small claims, and the ...
-
02 Apr 2026insurance-credibility v0.1.9 adds BMSEquilibriumSimulator — Lemaire NPV reporting thresholds, Liang 2-class equilibrium, and a frequency correction for the s...
-
02 Apr 2026Malandii & Uryasev's Biased Mean Quadrangle (arXiv:2603.26901) provides a linear-programming-based method for estimating E[Y]+x — a biased mean offset. For r...
-
02 Apr 2026insurance-monitoring v1.1.0 adds BAWSMonitor — bootstrap adaptive window selection for rolling VaR and ES estimates, replacing arbitrary lookback choices wit...
-
02 Apr 2026Every time you re-run conformal risk control calibration on a growing book, you are implicitly doing multiple testing. Hultberg et al. (2026) formalise the fix.
-
02 Apr 2026Laub/Pho/Wong's Actuarial Neural Additive Model has a genuine architectural insight in PWLCalibration monotonicity. It also depends on an unmaintained Tensor...
-
02 Apr 2026A runnable Python implementation of Goffard, Piette, and Peters (ASTIN Bulletin 2025): infer claim frequency and severity from competitor PCW quotes using AB...
-
01 Apr 2026Izbicki and Rodrigues (arXiv:2603.26611, March 2026) benchmark TabPFN-2.5, RealTabPFN-2.5 and TabICL-Quantiles as conditional density estimators across 39 da...
-
01 Apr 2026Yanez, Guillen and Nielsen (ASTIN Bulletin 2025) apply a bounded Bonus-Malus System not to claims but to telematics signals themselves, updating weekly. The ...
-
01 Apr 2026Zhang, Mao and Wang (arXiv:2603.14991, March 2026) prove a closed-form equivalent for worst-case quantile regression under Wasserstein distributional uncerta...
-
01 Apr 2026The FCA has flagged travel insurance for mental health conditions and contents insurance for social renters as supervisory priorities. Here is what ML can ge...
-
01 Apr 2026UK territory rating is mostly postcode-to-band lookup tables. That creates both actuarial and regulatory risk. We work through the spatial statistics toolbox...
-
01 Apr 2026Most retention models measure whether a customer lapses. surv-iTMLE measures when — and what your pricing intervention caused. We explain the estimand, why l...
-
01 Apr 2026Threshold selection is the Achilles' heel of extreme value theory in insurance pricing. Majumder and Richards (arXiv:2504.19994) eliminate it by blending a s...
-
01 Apr 2026Standard conformal prediction gives symmetric intervals calibrated on average. For right-skewed claim distributions, that average includes a lot of zero clai...
-
01 Apr 2026Medina-Olivares, Xia, Lessmann and Klein (arXiv:2603.26309, March 2026) build a semi-structured neural network model for mortgage delinquency transitions. Th...
-
01 Apr 2026Two rigorous frameworks for automated underwriting triage — SelectiveConformalRC (SCRC) controls expected loss on your auto-priced book; SCoRE controls total...
-
01 Apr 2026Roy, Singh, and Das (arXiv:2603.14841) build a 0-100 driver safety score by inverting a crash classifier and multiplying in condition-specific penalty factor...
-
01 Apr 2026A follow-up to our QPP introduction: the honest case for quantile-based loading (it works for heavy-tailed lines and low-frequency risks, it does not work be...
-
01 Apr 2026Protected NCD is widely misunderstood by consumers, and the product may not deliver the value it charges for. The Consumer Duty fair value test and the hunge...
-
01 Apr 2026Every UK pricing GLM pipeline that uses Lasso variable selection then reports Wald confidence intervals is producing coverage rates of 70–80% at a nominal 95...
-
01 Apr 2026PenalizedGLMInference in insurance-gam v0.5.0 implements Manna et al. arXiv:2410.01008: bias-corrected confidence intervals for Poisson, Gamma, and Tweedie G...
-
01 Apr 2026Standard conformal prediction gives valid coverage only when calibration and test data are exchangeable. For insurance models deployed for 12+ months — throu...
-
01 Apr 2026Two January 2026 arXiv preprints formalise what UK pricing teams have long intuited: observed claim frequency at high-NCD classes understates true frequency ...
-
01 Apr 2026Standard discrimination-free pricing applies to a single outcome. Income protection premiums are derived from a matrix of transition rates. Applying Lindholm...
-
01 Apr 2026An MGA launching on a UK PCW needs prices on day one with zero claims history. Here is the full architecture: market ABC as the prior, Bühlmann-Straub blendi...
-
01 Apr 2026Braun et al. (arXiv:2507.20941) replace the standard hyperrectangular joint prediction set with an ellipsoid built from a Mahalanobis nonconformity score. Fo...
-
01 Apr 2026LoBoostCP in insurance-conformal v1.0.0 implements Santos et al. arXiv:2602.22432 — local conformal prediction that uses the leaf structure of your existing ...
-
01 Apr 2026k-ary randomised response applies symmetric noise — wasteful and fairness-suboptimal. The Ghoukasian-Asoodeh optimal mechanism is asymmetric: minority groups...
-
01 Apr 2026A model can pass its age fairness audit and its gender fairness audit and still systematically overprice young women. This is fairness gerrymandering. We exp...
-
01 Apr 2026insurance-cv v0.3.0 adds SupportPointSplit (distributional train-test splitting via energy distance minimisation) and ChatterjeeSelector (nonlinear feature s...
-
01 Apr 2026insurance-monitoring v0.10.0 adds PricingDriftMonitor (Brauer/Menzel/Wüthrich arXiv:2510.04556) and CalibrationCUSUM (Franck et al. arXiv:2510.25573): formal...
-
01 Apr 2026The FCA's February 2026 insurance priorities report signals active supervisory review of AI pricing in Q1-Q2 2026. Under existing PRIN 2A, firms must already...
-
01 Apr 2026The FCA published its first annual insurance regulatory priorities document in February 2026. Here is everything a pricing actuary needs to do in response — ...
-
01 Apr 2026Solvency II requires age-based pricing. EU AI Act Article 21 lists age as a prohibited characteristic. Chouldechova's impossibility theorem shows you cannot ...
-
01 Apr 2026Age is a legally permitted rating factor under Solvency II. Age is also a protected characteristic under EU law. The EU AI Act imposes non-discrimination obl...
-
01 Apr 2026A UK MGA at launch has five routes to market data: competitor quote reverse-engineering, rate indices from Consumer Intelligence, ABI aggregate statistics, c...
-
01 Apr 2026EIOPA published its AI Governance Opinion (EIOPA-BoS-25-360) in August 2025. It names the actuarial function as responsible for AI controls, endorses SHAP an...
-
01 Apr 2026DML removes the omitted variable bias that makes naive GLM price elasticity estimates wrong by 20–80%. We explain why it works, show the two core insurance a...
-
01 Apr 2026insurance-gam v0.4.0 adds DebiasedGLM — the first Python implementation of debiased Lasso confidence intervals for Poisson, Gamma, and Tweedie GLMs. It corre...
-
01 Apr 2026Paolo Toccaceli's CRPS-Optimal Binning for Conformal Regression (arXiv:2603.22000) partitions the covariate space using dynamic programming to minimise LOO-C...
-
01 Apr 2026Verschuren (2021) showed that a Dutch insurer's home claim history predicts motor risk, and vice versa. The framework is technically clean. The UK structural...
-
01 Apr 2026Baradel (arXiv:2603.11258) extends continuous-time bootstrapping to Schnieper's model, separating IBNR (new claims) from IBNER (cost development of known cla...
-
01 Apr 2026Vadlamani et al. (ICLR 2025, arXiv:2505.16115) formalise fairness at the prediction-set level. A model can be statistically valid at 90% coverage while cover...
-
01 Apr 2026Burger (arXiv:2512.23602, Dec 2025) applies conformal prediction to insurance model monitoring, replacing PSI > 0.2 and A/E > 1.15 with thresholds that are c...
-
01 Apr 2026Standard conformal prediction fails with right-censored survival data because you never observe the true event time for censored policies. ConformalisedSurvi...
-
01 Apr 2026Standard conformal prediction breaks under instrumental variable regression — the calibration residuals are not exchangeable. Kato (arXiv:2603.25509, March 2...
-
01 Apr 2026PSI thresholds of 0.1 and 0.2 are industry convention, not statistical calibration. Burger (arXiv:2512.23602, Dec 2025) replaces them with conformal p-values...
-
01 Apr 2026Independent Lee-Carter models per cause-of-death produce forecasts that do not sum to total mortality — a coherence failure that flows directly into CI and L...
-
01 Apr 2026Kończal, Balcerek and Burnecki (arXiv:2512.22660, December 2025) systematically benchmark eight climate oscillation indices as ML features for predicting CAT...
-
01 Apr 2026Zhang, Guillen, Li, Li and Chen (arXiv:2509.02614) build a group-based zero-inflated Poisson model for ADAS near-miss event counts using 354 commercial drive...
-
01 Apr 2026ANAM (Laub, Pho, and Wong, NAAJ 2025) fits each rating factor as a neural subnetwork with hard monotonicity constraints, exposure offsets, and proper actuari...
March 2026
339 articles
-
31 Mar 2026Shin, Lee and Kang (arXiv:2512.03738, Dec 2025) provide the first finite-sample coverage guarantee for time-to-event prediction under covariate shift. Here i...
-
31 Mar 2026When you launch a new product with no claims history, you borrow from a related portfolio. Transfer learning formalises this. But the most-cited deep learnin...
-
31 Mar 2026Text embeddings can compress 13,000 business categories into 24 useful features and add 2-3% Gini lift on commercial lines. For UK personal motor FNOL they d...
-
31 Mar 2026Standard telematics pricing throws away most of the information in a trip by reducing it to 100+ scalar features. Two new papers from Toronto's Badescu group...
-
31 Mar 2026Tab-TRM sets the French MTPL benchmark at 23.589×10⁻² Poisson deviance, beating PIN ensemble by 0.3%. The linearisation result — Tab-TRM is approximately a s...
-
31 Mar 2026Naively applying Wasserstein barycenter corrections sequentially across multiple protected attributes is miscalibrated: the ECDF for attribute k was fitted o...
-
31 Mar 2026Lindholm, Richman, Tsanakas and Wüthrich (EJOR, January 2026) give us a scalar proxy discrimination measure with a property none of the existing methods shar...
-
31 Mar 2026Boonkrong et al. (MDPI Risks 14(3):57, March 2026) show that adding actuarial pricing features to a renewal classifier materially improves prediction — the i...
-
31 Mar 2026Li, Luo, Zhang, Huang and Jiang (IME 2025) combine telematics risk scoring with individual price sensitivity estimation and constrained discount allocation. ...
-
31 Mar 2026A January 2025 paper by Dong & Finlay (arXiv:2501.06404) combines a Variational Autoencoder with Proximal Policy Optimisation to dynamically adjust excess of...
-
31 Mar 2026The quantile premium principle maps a single number — your risk appetite parameter tau — to per-risk safety loadings. Zanzouri et al. (NAAJ 2025) shows QRNN ...
-
31 Mar 2026Thibodeau et al. build a multi-firm market simulator and demonstrate the collusion pathology: a cartel that excludes every income group equally passes standa...
-
31 Mar 2026Membership Inference Attacks are essentially uninformative on tabular insurance data — a finding from Zuo et al. (arXiv:2602.09288) with direct consequences ...
-
31 Mar 2026UK SME cyber is growing at 13% CAGR with 2.8% standalone penetration. The data is terrible, the commercial tools are six figures, and the best published freq...
-
31 Mar 2026Pearse & Bondell (arXiv:2510.06177, October 2025) derive a new Archimedean copula family from Cressie-Read phi-divergences. The generator table maps KL diver...
-
31 Mar 2026ACI satisfies its marginal coverage guarantee while producing months of invalid intervals after a claims inflation shock. A new paper proves the minimax-opti...
-
31 Mar 2026Boucher & Coulibaly (arXiv:2502.11788) prove that for Poisson frequency models, log-exposure offset and exposure-weighted ratio approaches are identical. The...
-
31 Mar 2026Policyholders with good NCD rationally choose not to report small claims. Your frequency model is trained on that suppressed data. Two January 2026 papers fo...
-
31 Mar 2026Denuit, Michaelides & Trufin (arXiv:2603.16317) prove that autocalibration and group fairness are mathematically equivalent. A GBM that is well-calibrated ov...
-
31 Mar 2026Lim, Xu & Zhou (arXiv:2602.04791) show that multi-state models decompose into independent Poisson GLMs — one per transition — making every single-period fair...
-
31 Mar 2026The Goffard/Piette/Peters ABC-SMC method infers risk parameters from competitor quotes with no claims history. No Python implementation exists - only IsoPric...
-
31 Mar 2026LLMs encode societal stereotypes. When you use one to generate rating features, those stereotypes enter your pricing model. The insurer is responsible. Here ...
-
31 Mar 2026Three published frameworks use LLMs to generate tabular features and beat classical search tools on generic benchmarks. None has been tested on an actuarial ...
-
31 Mar 2026Mo et al. (arXiv:2602.16836, Feb 2026) fine-tuned DeepSeek-R1-Distill-Llama-8B on 2 million automotive warranty claims. The LoRA-tuned model hit 81.5% struct...
-
31 Mar 2026Laghuvarapu, Deb and Sun (arXiv:2603.26415, March 2026) replace per-test-point density ratio estimation with bounded QP weights on the calibration set. Here ...
-
31 Mar 2026Miao & Pesenti (arXiv:2603.16720) derive the discrimination-insensitive pricing measure from first principles: find the nearest probability measure Q to the ...
-
31 Mar 2026A one-shot OLS regression on individual claims — using the adjuster's case reserve history as input — outperforms chain-ladder on liability data. Richman and...
-
31 Mar 2026The new InflationDecomposer in insurance-trend v0.1.5 uses the Harvey structural time series model to separate persistent cost inflation from cyclical effect...
-
31 Mar 2026Brauer, Menzel & Wüthrich (arXiv:2510.04556) give us two things we have been missing: a formal hypothesis test for Gini drift and a Murphy score decompositio...
-
31 Mar 2026arXiv:2510.24601 reviewed 143 papers across 430 datasets and found no consistent accuracy advantage for neural networks over GAMs on tabular data by 2024. Wh...
-
31 Mar 2026Focal loss is a clever idea from computer vision that does not translate well to tabular insurance fraud data. AUC=0.63 from a three-stage focal loss neural ...
-
31 Mar 2026FL solves the same variance-reduction problem as Bühlmann-Straub — but iteratively, with communication overhead, and without actuarial precedent. For UK pers...
-
31 Mar 2026EBMs achieve near-XGBoost predictive performance on insurance claims data while remaining fully interpretable by design — no post-hoc SHAP required. We show ...
-
31 Mar 2026Ignoring policy limits and IBNR censoring when fitting tail distributions biases your tail index by ~15%. For UK motor TPBI with xi in the 0.50–0.67 range, e...
-
31 Mar 2026Sun, Xie & Zhang (arXiv:2503.11375) combine parallel trends and synthetic control into a single estimator that remains consistent if either assumption holds....
-
31 Mar 2026GLM prediction intervals targeting 95% coverage achieved 57.8% actual coverage on real personal injury data. Conformal prediction fixes this without distribu...
-
31 Mar 2026Zhang, Liu and Shi (arXiv:2504.11775, 2025) extend discrimination-free pricing to the case where you only have a noisy privatised version of the protected at...
-
31 Mar 2026DP-CTGAN produces near-random output at epsilon=1 on datasets under 50K rows — which is most insurance portfolios. AIM via smartnoise-synth is the correct to...
-
31 Mar 2026Dey (arXiv:2601.11949) builds a two-step pipeline — MLP for precipitation-to-claims, Gumbel copula for climate model uncertainty — that is methodologically s...
-
31 Mar 2026If you are using XGBoost's monotone_constraints or applying isotonic regression post-hoc, you may already have everything you need. Or you may not. The answe...
-
31 Mar 2026Parametric Tweedie intervals over-cover low-risk policies and under-cover the high-risk tail. Conformal prediction fixes this with a finite-sample guarantee ...
-
31 Mar 2026Conformal Prediction with Change Points: When Your Coverage Guarantee Breaks and What to Do About ItConformal prediction with change points (CPTC, arXiv:2509.02844, NeurIPS 2025) extends adaptive conformal inference to detect structural breaks and reset cov...
-
31 Mar 2026Hinder et al. (arXiv:2602.19790, ESANN 2026) introduce bootstrap conformal p-values for identifying which individual observations are affected by drift. We e...
-
31 Mar 2026Conformal prediction gives valid marginal coverage but says nothing about conditional coverage — your intervals can fail for young drivers or flood-zone prop...
-
31 Mar 2026Marginal coverage guarantees say nothing about which policyholders are being undercovered. CVI decomposes conditional coverage into undercoverage risk and ov...
-
31 Mar 2026D-calibration and ICI are mathematically invalid for competing-risks models. If F_k(inf|x) < 1 — which is always true for lapse, claim, and MTA competing cau...
-
31 Mar 2026Fitting one aggregate trend to UK motor claims 2019–2024 embeds a single implicit decay rate across parts shortage, labour shortage, and social inflation — c...
-
31 Mar 2026Wang, Shi, and Cao (NAAJ 2025) propose clustering NN-encoded residuals under a spatial contiguity constraint to produce hard territory zones. Here is how tha...
-
31 Mar 2026EA NaFRA is open, 2m resolution, and free. So is the EPC register. So are OS building footprints. A UK pricing actuary who has actually tried to use them exp...
-
31 Mar 2026Bühlmann-Straub pulls every segment toward the same grand mean. When that mean is wrong for your segment, the correction makes things worse. We explain when ...
-
31 Mar 2026Pittarello, Hiabu, and Villegas (NAAJ 2025) showed that chain ladder is the age-only special case of an Age-Period-Cohort model borrowed from demography. We ...
-
31 Mar 2026Laub, Pho and Wong's ANAM gives each rating factor its own neural network sub-model, sums them like a GAM, and adds three things a generic NAM cannot do: har...
-
30 Mar 2026Your A/E ratio is drifting. Is it the portfolio mix, or has the underlying loss mechanism changed? The answer determines whether you update the intercept or ...
-
30 Mar 2026Every UK fairness audit that uses LSOA ethnicity percentages as its protected attribute is reporting a lower bound on bias, not the true figure. Here is what...
-
30 Mar 2026Standard UK practice for POT threshold selection is to pick a round number and hope. Two recent papers — EQD (Murphy et al. 2024) and BMA (Jessup et al. 2025...
-
28 Mar 2026The first pip-installable implementation we know of for So & Valdez (2024) Scenario 2: a two-stage CatBoost model that separates structural zero probability ...
-
28 Mar 2026insurance-distributional v0.3.0 ships ZeroInflatedTweedieGBM — the first open-source implementation of So & Valdez (2024) Scenario 2. When standard Tweedie g...
-
28 Mar 2026We fit WhittakerHendersonPoisson to driver age frequencies from 677K French MTPL policies. The Poisson smoother handles count data correctly, REML selects la...
-
28 Mar 2026Why Tweedie GLM is the standard for aggregate loss modelling in insurance, with a complete Python example covering power parameter selection, exposure offset...
-
28 Mar 2026An honest assessment of where tabular foundation models stand in March 2026 — what the benchmarks actually show, what's missing for insurance pricing, and wh...
-
28 Mar 2026Three-way benchmark on 677K French motor policies. TabPFN cannot handle log-exposure offsets — the structural limitation that makes it unviable for bread-and...
-
28 Mar 2026Deep learning survival models underperform Cox regression on tabular insurance data. Cure models are the real story post-GIPP. Here is what the research says...
-
28 Mar 2026UK motor average claim costs reached a record £5,300 in Q4 2024. But applying a flat 8% trend assumption treats structural and cyclical inflation identically...
-
28 Mar 2026How to produce a full IBNR distribution in Python using the Mack method and Bootstrap ODP sampling. Covers analytical standard errors, 5,000-simulation boots...
-
28 Mar 2026Your GLM or GBM was trained on policyholders who chose to buy from you at your price. That is not a random sample of the market. The mechanism, what it does ...
-
28 Mar 2026A practitioner-oriented deep dive on applying reinforcement learning and contextual bandits to PCW margin optimisation for UK personal lines. Two serious pap...
-
28 Mar 2026XGBoostLSS, LightGBMLSS, NGBoost, and PGBM can all output a full conditional distribution rather than a point prediction. The Chevalier & Côté benchmark (EAJ...
-
28 Mar 2026In January 2026 the PRA named AI as an insurance supervisory theme for the first time. The FCA published a bias research note in December 2024. The FRC updat...
-
28 Mar 2026Theory proves standard XL layers are optimal under VaR. Yet most pricing teams set retentions by trial and error. We review the academic results, the commerc...
-
28 Mar 2026When does it make sense to reach beyond chain ladder and bootstrap ODP for neural reserving methods? We compare DeepTriangle, individual RNN approaches, and ...
-
28 Mar 2026End-to-end motor insurance pricing in Python using the French MTPL dataset. Frequency-severity GLMs, exposure offsets, coefficient interpretation, validation...
-
28 Mar 2026Most governance tooling is tested on toy examples with clean DGPs and inflated Gini coefficients. We ran the full insurance-governance validation suite on 67...
-
28 Mar 2026Gamma GLMs fit a single mode to severity data that often has two or three. Mixture Density Networks output the full conditional distribution — mixing weights...
-
28 Mar 2026The standard rate adequacy workflow — earned premium at current rate level, ultimate losses, trend to future period, expense load, indicated rate change — bu...
-
28 Mar 2026Richman and Wüthrich's March 2026 paper shows linear regression with projection-to-ultimate factors closes 44% of the gap over chain ladder — and neural netw...
-
28 Mar 2026Kolmogorov-Arnold Networks replace fixed activations with learnable splines on edges, letting the model discover its own functional forms. Here is what that ...
-
28 Mar 2026Raw loss ratios by age band are noisy. A 5-year moving average introduces boundary bias and requires a judgment call you cannot defend in an IFRS 17 review. ...
-
28 Mar 2026The GBM sits in a notebook outperforming the production GLM. This tutorial shows how to extract multiplicative rating relativities from a CatBoost Poisson mo...
-
28 Mar 2026Point estimates from pricing models are incomplete. This tutorial shows how to add distribution-free prediction intervals to a CatBoost Tweedie model using i...
-
28 Mar 2026Fit a Tweedie GLM to insurance data in Python, extract multiplicative rating relativities, and visualise factor curves — with working code throughout.
-
28 Mar 2026The FCA published its final premium finance market study on 3 February 2026. No APR cap was imposed. That does not mean your book is clean. Here is what chan...
-
28 Mar 2026The FCA's 2026 access priorities are not a compliance problem — they are an actuarial evidence problem. The question is whether the data exists to justify wh...
-
28 Mar 2026Single-objective fairness constraints force a binary choice. NSGA-II finds the full tradeoff surface, so governance committees can make an explicit, document...
-
28 Mar 2026We published a post three weeks ago treating all insurance pricing models as potentially high-risk under the EU AI Act. After reading the primary sources, we...
-
28 Mar 2026Most industry guidance on the EU AI Act and insurance pricing is wrong in ways that matter. Motor and property pricing is not high-risk AI. Traditional GLMs ...
-
28 Mar 2026A step-by-step guide to building an EU AI Act conformity assessment for an insurance pricing model. Covers risk classification, Article 11 technical document...
-
28 Mar 2026Article 13 of the EU AI Act is not about SHAP values. It is about deployer-facing documentation — what the underwriter or product team needs to interpret and...
-
28 Mar 2026Most UK motor insurers think they know their price elasticity. They are probably wrong by a factor of 3–5, in the direction that makes them systematically mi...
-
28 Mar 2026A practitioner's guide to dynamic pricing in UK insurance: what GIPP actually permits, why your elasticity model likely has a 3-5x bias, and an honest assess...
-
28 Mar 2026We benchmarked Whittaker-Henderson against raw rates and a 5-point weighted moving average on a synthetic UK motor driver age curve with known truth. W-H red...
-
28 Mar 2026The standard UK motor pricing formula multiplies E[N] by E[S] and assumes independence. On a 15,000-policy benchmark with planted omega=3.5, that assumption ...
-
28 Mar 2026PSI detects covariate shift but not rank collapse. On a synthetic UK motor book where a new risk factor emerges post-deployment, PSI stays GREEN while Gini d...
-
28 Mar 2026Manual Spearman correlation missed postcode as an ethnicity proxy in 100% of 50 benchmark runs. CatBoost proxy R-squared caught it in 100% of runs. The diffe...
-
28 Mar 2026On a UK motor DGP with a monotone young-driver requirement, unconstrained EBM violates monotonicity in 31% of runs. Constrained EBM matches GLM monotonicity ...
-
28 Mar 2026HMM-derived driving state features improve Gini by 5–10 percentage points over raw trip averages on a state-structured DGP. The reason is temporal: the HMM k...
-
28 Mar 2026We benchmarked constrained portfolio optimisation against a uniform +7% rate change on a 2,000-policy UK motor book. The optimiser achieved the same GWP targ...
-
28 Mar 2026We benchmarked Bühlmann-Straub credibility against raw experience and manual Z-factors on a 30-segment synthetic UK motor fleet book with a known DGP. On thi...
-
28 Mar 2026REML-selected lambda beats manual tuning on a 63-band age curve benchmark: 22% lower MSE on thin tail bands, zero analyst discretion, and principled credible...
-
28 Mar 2026Miao & Pesenti (arXiv:2603.16720) derive discrimination-insensitive premiums by finding the probability measure nearest to the real-world measure in KL-diver...
-
28 Mar 2026Build loss development triangles, calculate IBNR reserves, and plot development patterns using Python and the chainladder library.
-
28 Mar 2026BonusMalus is built from past claims — a naive regression conflates the causal effect with selection. We ran Double Machine Learning on 677K freMTPL2 policie...
-
28 Mar 2026Build a working Solvency II SCR estimate from scratch using compound distributions and Monte Carlo simulation. Poisson/NegBin frequency, lognormal severity, ...
-
28 Mar 2026A practical Python walkthrough of the burning cost method for pricing excess of loss reinsurance treaties — loss trending, development, pure rate calculation...
-
28 Mar 2026We ran Bühlmann-Straub credibility on the freMTPL2freq dataset — 677K French MTPL policies, 22 regions — and quantified how much thin regions get pulled towa...
-
28 Mar 2026H2O, FLAML, and AutoGluon are genuinely useful tools. None of them handle the log(exposure) offset that makes insurance frequency modelling work. Here is an ...
-
27 Mar 2026A definitive survey of open-source Python tools for insurance pricing in 2026. General-purpose ML libraries, specialist actuarial packages, the Burning Cost ...
-
27 Mar 2026Denuit, Michaelides and Trufin (March 2026) unify autocalibration and non-discrimination into a single actuarial test. If your model fails it, you have a pri...
-
27 Mar 2026We fitted a Poisson GLM on the first third of freMTPL2 (677k French motor policies) and monitored it across two later temporal segments without refitting. PS...
-
27 Mar 2026We ran insurance-fairness against ausprivauto0405 — a real Australian motor dataset with an explicit Gender field. Here is what FairnessAudit, Multicalibrati...
-
27 Mar 2026Ciganovic et al. (March 2026) show that standard DML cross-fitting leaks future information when your data is a time series. Their fix — Reverse Cross-Fittin...
-
27 Mar 2026We planted three simultaneous model failures in a 50,000-policy UK motor book. The aggregate A/E never triggered. The library detected the first problem afte...
-
27 Mar 2026Sesia & Favaro's March 2026 survey of conformal prediction is the clearest account yet of what finite-sample distribution-free guarantees mean - and why th...
-
26 Mar 2026Lee, Badescu, and Lin (2026) replace ad-hoc event counts with a principled actuarial risk index: MODWT decomposes the acceleration signal, a Gaussian-Uniform...
-
26 Mar 2026A complete Python tutorial for building a Tweedie GLM for insurance pricing: synthetic motor data, statsmodels, exposure offset, interpreting the p parameter...
-
26 Mar 2026The Civil Liability Act 2018 split UK TPBI into two structurally different populations. Standard frequency-severity models treat them as one. Here is why tha...
-
26 Mar 2026Most UK insurers fit a logistic regression on PCW quote data and call it a demand model. It is biased in at least three distinct ways. Here is the causal str...
-
26 Mar 2026Brehmer & Strokorb (2019) proved that no proper scoring rule applied to raw data can discriminate tail indices. Bladt & Øhlenschlæger (arXiv:2603.24122) fix ...
-
26 Mar 2026A stochastic SIR model calibrated to LockBit ransomware data shows why treating cyber losses as independent events badly underestimates portfolio-level risk.
-
26 Mar 2026Balzer and Benlahlou (arXiv:2603.14543) extend gradient boosting to spatial panel data. Here is what it does, how it compares to BYM2 and Blier-Wong, and whe...
-
26 Mar 2026Balzer and Benlahlou's spatial GBM uses GMM pre-estimation and a Cochrane-Orcutt transformation to handle spatial autocorrelation in gradient boosting. It is...
-
26 Mar 2026A new paper from ETH Zürich (arXiv:2603.18653) frames the conversion from technical price to commercial premium as a Multiple-Choice Knapsack Problem. Under ...
-
26 Mar 2026Richman-Wüthrich's one-shot PtU reserving paper (arXiv:2603.11660) ships with R code only. We map the algorithm to Python, explain the censored-claims exposu...
-
26 Mar 2026A new paper (Xin, Hooker, Huang 2026) shows that BIFSG proxy race distorts regression-based fairness audits in two distinct mechanisms — and the direction of...
-
26 Mar 2026A new paper combines panel fixed effects, double machine learning, and instrumental variables. The headline result is not the estimator — it's that ML covari...
-
26 Mar 2026SAM2, GroundingDINO, YOLOv8, Roboflow Universe — the open-source CV stack for building analysis exists. This is an honest account of what it can do, what tra...
-
26 Mar 2026How to turn insurance claims descriptions into GLM features using sentence-transformer embeddings and PCA. What Troxler & Schelldorfer (2024, BAJ) actually s...
-
26 Mar 2026Python has no equivalent of R's msm package for continuous-time multi-state modelling of claims. We explain the mathematics, show why a Poisson GLM substitut...
-
26 Mar 2026Samuel Clark's MDMx organises the Human Mortality Database as a four-way tensor and applies Tucker decomposition to produce structurally coherent mortality m...
-
26 Mar 2026The LRE metric converts Pearson correlation improvement into expected loss ratio change in basis points. Here is how to use it, and where it breaks.
-
26 Mar 2026LLMs can extract structured data from surveys, flag non-standard construction in loss adjuster notes, and rate categorical variables at scale. They cannot re...
-
26 Mar 2026Moriah et al. (2026) run a sequential model-building exercise on a French home insurance portfolio to measure what each data layer — hydrological zoning, rai...
-
26 Mar 2026Two deregulatory signals in the FCA's February 2026 insurance priorities document. The SMCR one is almost entirely administrative. The Consumer Duty one is r...
-
26 Mar 2026FCA PS26/2 (March 2026) creates mandatory incident reporting and material third party registers for all authorised insurers. Every pricing actuary who owns a...
-
26 Mar 2026The FCA's Mills Review is not another compliance checklist — it is the regulator thinking aloud about how UK financial services might look in 2030. Pricing t...
-
26 Mar 2026Why the standard flat EV surcharge is wrong in two directions simultaneously, what the claims data actually shows, and how to build a severity model that han...
-
26 Mar 2026How to estimate a causally identified price elasticity from PCW quote data in Python, using commercial loading variation as an instrument and CatBoost nuisan...
-
26 Mar 2026Parametric Tweedie intervals undercover high-risk policies by 10–15 percentage points. We tested conformal prediction on 50,000 UK motor policies to find out...
-
26 Mar 2026When you have fewer than 5,000 policies in a segment, should you use Bühlmann-Straub credibility or a GBM with transfer learning? The answer depends on wheth...
-
26 Mar 2026OEM APIs, smartphone SDKs, and charging data — what connected-car data sources UK pricing teams can actually access, and how to turn them into rating factors.
-
26 Mar 2026Conformal prediction and the parametric bootstrap both produce prediction intervals for insurance pricing models. They answer different questions, have diffe...
-
26 Mar 2026Conformal prediction gives finite-sample valid 99.5% risk bounds for individual policies — useful for premium risk SCR validation and model validation consis...
-
26 Mar 2026Cape Analytics has 700k+ claim validations and a 400% loss ratio differential between best and worst roof condition. The UK market has none of that infrastru...
-
26 Mar 2026Li and Castro-Camilo (arXiv:2603.23309, March 2026) unify inverse probability weighting and extreme value extrapolation in a single estimating equation. Here...
-
26 Mar 2026Fang, Tan, Pipping, and Hooker (AISTATS 2026) show that replacing additive boosting with a moving-average update makes EBMs converge to kernel ridge regressi...
-
25 Mar 2026A structured decision framework for choosing between conformal prediction, distributional GBM, Bühlmann-Straub credibility, GLM bootstrap, and GAM uncertaint...
-
25 Mar 2026EY forecasts a 111% net combined ratio for UK motor in 2026. WTW documents a 13% annual premium fall. Here is what the data shows and what pricing teams shou...
-
25 Mar 2026The adverse selection trap in opt-in UBI: why telematics discounts attract the risks you least want to retain, and what to do about it.
-
25 Mar 2026TabPFN v2 (Nature 637:319–326, 2025) does zero-shot prediction on datasets up to 10K rows. Here is what that actually means for the pricing segments where yo...
-
25 Mar 2026Tab-TRM (arXiv:2601.07675) is a 14,820-parameter recursive model that beats CatBoost on French MTPL while connecting to GLM theory. We explain the architectu...
-
25 Mar 2026Havrylenko et al. (2025) show that MICE with random forests outperforms CTGAN and VAEs on the freMTPL2 benchmark. We explain why it works, where it fails, an...
-
25 Mar 2026Applying CANN + NID to severity (Gamma) GLMs. Why the signal is weaker than frequency, what configuration changes are needed, and when a severity interaction...
-
25 Mar 2026Avanzi, Richman, and Wüthrich reformulate individual claims reserving as a Markov Decision Process. We explain why it matters, what it actually does, and whe...
-
25 Mar 2026How to combine lapse hazard models with causal price elasticity under PS21/5 constraints for UK motor and home renewal pricing.
-
25 Mar 2026The complete PS21/5 compliance workflow: CATE estimation with insurance-causal, ENBP-constrained optimisation with insurance-optimise, fairness audit with in...
-
25 Mar 2026Chevalier & Côté (EAJ 2025) benchmark nine GBM variants on five insurance datasets. We read it so you don't have to, then show where insurance-distributional...
-
25 Mar 2026UK GDPR constrains what pricing data you can share across entities. Federated learning and differential privacy offer a way around the constraint — but only ...
-
25 Mar 2026Zero internal claims data is not a reason to guess blindly. Here is a structured sequence of five approaches — from Bühlmann-Straub credibility priors throug...
-
25 Mar 2026FCA MS24/2 (February 2026) means pricing teams now own the APR question. Here is how to treat it as a pricing problem — with the same tools used for the insu...
-
25 Mar 2026How UK home insurers should model physical climate risk: UKCP18 projections, Flood Re's 2039 exit, ABI claims data, and practical code using insurance-whitta...
-
25 Mar 2026The FCA has explicitly flagged pet insurance for monitoring in its 2026 regulatory priorities. FOS complaint upheld rates hit 52% in Q1 2025 — the highest of...
-
25 Mar 2026SOA and CAS research from late 2025 has sharpened the methods for calibrating parametric triggers and quantifying basis risk. Here is what that means in prac...
-
25 Mar 2026arXiv:2603.11660 proposes direct projection-to-ultimate on individual claims data. The honest finding: linear regression on claim status and incurred already...
-
25 Mar 2026NSGA-II finds the non-dominated pricing strategies across accuracy, group fairness, and counterfactual fairness simultaneously. TOPSIS turns that front into ...
-
25 Mar 2026How to detect when a motor book has hit the floor of its underwriting cycle — using PSI on new business mix, segment-level A/E, Gini stability, and mSPRT to ...
-
25 Mar 2026Modern Insurance Pricing with Python and Databricks - all 12 modules, free, on GitHub. GLMs through causal elasticity, fairness auditing, spatial BYM2 territ...
-
25 Mar 2026A hands-on tutorial for the RateChangeEvaluator in insurance-causal v0.6.0. DiD when you have a control group. ITS when you don't. Real code, real API.
-
25 Mar 2026Goffard, Piette, and Peters (ASTIN Bulletin 2025) show how to calibrate insurance rates using competitor premiums and no internal claims data — using ABC and...
-
25 Mar 2026A 5pp Gini improvement means nothing to a CFO. The Loss Ratio Error framework from arXiv:2512.03242 converts model correlation into expected loss ratio — and...
-
25 Mar 2026Scenario modelling for UK motor bodily injury claims under whiplash reform uncertainty — Taylor separation, severity distributions, and conformal prediction ...
-
25 Mar 2026Build a double-lift chart to compare GLM vs GBM predictions. Bin by prediction ratio, compute A/E per decile, plot with matplotlib. Standard tool for pricing...
-
25 Mar 2026The average treatment effect hides a 5x spread in price elasticity across a UK motor book. GATES, CLAN, and RATE tell you the size, who's who, and whether th...
-
25 Mar 2026The first Python implementation of the asymptotic Gini drift test from Brauer et al. (2025). A proper z-test for ranking degradation — not a heuristic, not a...
-
25 Mar 2026Two approaches to prediction intervals for insurance severity: distributional GAMLSS (insurance-distributional-glm) vs distribution-free conformal (insurance...
-
25 Mar 2026The FCA's interim report on MS24/1 landed in January 2026 with a Q3 2026 final report expected. Here is what pure protection pricing teams need to build befo...
-
25 Mar 2026FCA EP25/2 published July 2025. Expected claim costs per home policy up 49% from £92 to £138. Average inception premium up only 5%. The data says insurers ab...
-
25 Mar 2026FCA Consumer Duty PRIN 2A requires insurers to tell policyholders what they can change to get a better outcome. Most pricing teams have not built this. insur...
-
25 Mar 2026The FCA has committed to evaluating how UK insurers use AI in pricing, underwriting, and claims. No new AI-specific rulebook is coming — but the regulator no...
-
25 Mar 2026The FCA replaced 40+ portfolio letters with a single annual priorities document on 24 February 2026. Here is what it says, what it means for pricing teams, a...
-
25 Mar 2026A February 2026 paper from Lim, Xu, and Zhou cracks open the problem of fair pricing in multi-state insurance products — the ones that matter most for Consum...
-
25 Mar 2026Three recent papers on EVT and ML — from generalisation bounds for tail learning to Bayesian nonparametric splicing — and what they actually imply for UK sev...
-
25 Mar 2026A covariate that predicts mean severity well may tell you almost nothing about your 99th percentile claims. Here is how to identify which rating factors actu...
-
25 Mar 2026Life and health insurance pricing models are high-risk AI under Annex III of the EU AI Act. General insurance (motor, home) may also be in scope but the clas...
-
25 Mar 2026BCG's 2025 analysis puts embedded insurance at 30% CAGR. The pricing architecture question is not whether to do it - it's whether your model can answer in un...
-
25 Mar 2026We ran Double Machine Learning against a naive GLM on a 50,000-policy UK motor telematics book. The GLM overestimated the treatment effect by 50–90%. Here is...
-
25 Mar 2026insurance-fairness v0.6.3 ships DiscriminationInsensitiveReweighter. Here's why dropping the protected column doesn't work, how propensity-based reweighting ...
-
25 Mar 2026The FCA's first full outcomes monitoring year under Consumer Duty is under way. Here is what data to collect, what the regulator actually wants to see in MI,...
-
25 Mar 2026EP25/2 (the FCA's evaluation of GIPP price-walking remedies) flags ongoing fair value supervision in motor and home. No single technical checklist exists for...
-
25 Mar 2026How solvency_capital_range() produces model-free 99.5% SCR bounds, how SCRReport produces the coverage validation table for regulatory submission, and why in...
-
25 Mar 2026Fleet motor, property, and liability pricing in Python. Covers Bühlmann-Straub credibility for fleet schemes, GPD large loss loading, MBBEFD ILF tables, and ...
-
25 Mar 2026Why GLM coefficients aren't causal effects, and how to fix that using insurance-causal: DML with CatBoost nuisances, causal forests for heterogeneous treatme...
-
25 Mar 2026A clean Python tutorial for the most-cited neural network architecture in actuarial pricing: the Combined Actuarial Neural Network (Schelldorfer & Wüthrich, ...
-
25 Mar 2026Fiedler & Lucia's Bayesian Last Layer gives you calibrated posterior uncertainty from a neural network at near-zero additional cost. Here is what it does, wh...
-
25 Mar 2026Bühlmann-Straub credibility breaks when your hierarchy has more than two levels or when the random effects interact with pricing factors. Here is when to upg...
-
25 Mar 2026arXiv:2504.16592 formalises what pricing teams have been quietly observing for years: autonomous pricing algorithms can converge to supra-competitive prices ...
-
25 Mar 2026Laub, Pho and Wong's ANAM paper enforces smoothness and monotonicity architecturally, not as penalties. Here is what the mechanism actually is, why it matter...
-
24 Mar 2026Standard Tweedie GLMs handle zeros implicitly. When that implicit handling breaks — specialty lines, niche segments, specific peril models — you need ZIP or ...
-
24 Mar 2026How to implement walk-forward cross-validation for insurance GLMs in Python using insurance-cv. Covers IBNR buffers, fold design, and a full worked example o...
-
24 Mar 2026The Tweedie GLM and frequency-severity split both model pure premium. They are not interchangeable. Here is how to decide which one you actually need.
-
24 Mar 2026Every insurance team checks their champion/challenger results monthly. Every month you look, you inflate the false positive rate. Here is how to do it correc...
-
24 Mar 2026Monthly peeking at champion/challenger results with a t-test inflates your false positive rate to ~25%. The mixture SPRT (Johari et al. 2022) is an e-process...
-
24 Mar 2026Definitive Python benchmark: Poisson GLM vs XGBoost vs CatBoost vs LightGBM for insurance frequency modelling on freMTPL2. Poisson deviance, Gini coefficient...
-
24 Mar 2026PS25/21 abolished the mandatory 12-month product review cycle in December 2025. Harm-proportionate review cadence is now the requirement. Here is what that m...
-
24 Mar 2026How the Ogden discount rate and Periodical Payment Orders change the maths of large BI pricing in the UK — with Python code to calculate lump sum equivalents...
-
24 Mar 2026UK motor bodily injury severity has outrun CPI since 2022. This post implements a multiplicative severity separation model and Whittaker-Henderson smoothing ...
-
24 Mar 2026Migrating from Emblem to Python for insurance GLM pricing: what changes in workflow, what gets easier, what gets harder, and what the transition actually loo...
-
24 Mar 2026What large language models can genuinely contribute to insurance pricing feature engineering — text embeddings, zero-shot classification, synthetic features ...
-
24 Mar 2026Step-by-step: extract CatBoost factor tables with shap-relativities and write a clean Excel file with openpyxl. Formatted output ready to paste into Radar or...
-
24 Mar 2026Using causal forests and GATES/CLAN/RATE inference to find which customers respond most to a price or discount change — not just the average effect.
-
24 Mar 2026The Python equivalent of the IFoA MLR Working Party's R tutorial: Poisson GLM baseline, EBM GAM, and CatBoost GBM on UK motor data, with the full pipeline fr...
-
24 Mar 2026A practical statsmodels tutorial for pricing actuaries: Poisson frequency model with exposure offset, Gamma severity model, overdispersion tests, factor tabl...
-
24 Mar 2026Benchmark results on a known-DGP synthetic UK motor book. EBM beats the GLM by 12.6 Gini points (0.455 vs 0.329). But the deviance number is misleading. We e...
-
24 Mar 2026Benchmark results on a known-DGP synthetic UK motor fleet. HMM state fractions deliver 5–10pp Gini lift over simple aggregates. State classification recovers...
-
24 Mar 2026Honest benchmark: does fitting a surrogate GLM on CatBoost pseudo-predictions recover more discriminatory power than a direct GLM? We test it on 30,000 synth...
-
24 Mar 2026insurance-deploy provides the champion/challenger infrastructure, audit trail, and ICOBS 6B compliance tooling that MLflow does not. Here is how to use it.
-
24 Mar 2026ConformalisedQuantileRegression in insurance-conformal v0.6.2 gives you statistically guaranteed prediction intervals that are wide for high-risk segments an...
-
24 Mar 2026Extract the calendar-year inflation component from a claims development triangle using Taylor's two-factor separation. Python from scratch, then connect to s...
-
24 Mar 2026CPI-adjusting your historical claims data before fitting a pricing model introduces systematic bias. How to apply line-specific inflation indices for motor a...
-
23 Mar 2026A practical comparison of Python and R for UK personal lines insurance pricing — data wrangling, GLMs, GBMs, deployment, and Databricks. Honest about where R...
-
23 Mar 2026One-way analysis in Python for pricing actuaries: pandas from scratch, credibility-weighted confidence intervals, thin cell handling, GBM shortcuts.
-
23 Mar 2026A UK motor frequency model drifts after an upstream vehicle group reclassification. We show how insurance-monitoring's PSI, A/E ratios, and Gini drift test c...
-
23 Mar 2026Reproduce an Emblem frequency-severity GLM in Python: factor tables, one-way plots, deviance residuals, and lift charts using statsmodels, CatBoost, and Polars.
-
23 Mar 2026Which GLM assumptions actually matter for insurance pricing, which ones you routinely violate without consequence, and the diagnostics worth running before s...
-
23 Mar 2026A practical walkthrough for pricing analysts: use insurance-causal for causal inference, insurance-conformal for prediction intervals, and insurance-monitori...
-
23 Mar 2026Exposure-weighted Gini for insurance pricing: correct formula, Python implementation, and why ignoring exposure distorts motor model governance.
-
23 Mar 2026Benchmark results on a known-DGP synthetic UK motor age curve. REML recovers the true frequency well in the data-rich middle. The tails are a different story...
-
23 Mar 2026Aggregate A/E at 0.94 looks fine. The model has been mispricing under-25s for eight months. Benchmark results on a synthetic UK motor book with three planted...
-
23 Mar 2026We ran the benchmarks. On a synthetic UK motor book with nonlinear confounding, naive logistic GLM overestimates the telematics treatment effect by 50–90%. D...
-
23 Mar 2026Benchmark results on a known-DGP synthetic motor book. Conformal hits 90% across all deciles. Parametric Tweedie under-covers the top decile by 10–15pp. Numb...
-
23 Mar 2026Benchmark results on 100 synthetic schemes with known true loss rates. Credibility blending reduces MSE by 25–35% vs the best naive alternative. Numbers, not...
-
23 Mar 2026A practical comparison of CatBoost and XGBoost for UK personal lines insurance pricing — categorical handling, Tweedie support, and why we default to CatBoost.
-
23 Mar 2026Build a burning cost model in Python: frequency-severity split, exposure offsets, large loss capping, IBNR adjustment, and combined pure premium for UK pricing.
-
22 Mar 2026Three Ways to Detect Proxy Discrimination in Your Pricing Model: When They Agree and When They Don'tMutual information, proxy R-squared, and SHAP proxy scores all flag proxy discrimination but catch different things. A practical guide to interpreting confli...
-
22 Mar 2026sklearn's TweedieRegressor is a well-engineered GLM. It fits a fixed-power Tweedie model correctly. The problem is that insurance pricing needs per-risk vari...
-
22 Mar 202620 short code recipes for common insurance pricing tasks. Each recipe uses a real API from one of our open-source libraries. Copy and adapt.
-
22 Mar 2026Tutorial on monitoring insurance pricing models using actuarial KPIs. Gini tracking, segmented A/E, double-lift for champion/challenger. Why generic drift to...
-
22 Mar 2026Evidently is excellent for generic ML monitoring. It doesn't do exposure-weighted PSI, Poisson A/E ratios, Gini drift testing, or anytime-valid sequential te...
-
22 Mar 2026GLM insurance Python for UK pricing actuaries: exposure handling, Consumer Duty, frequency-severity split, and Emblem/Radar deployment with glum.
-
22 Mar 2026Under Consumer Duty and the Equality Act 2010, non-life insurers must test whether rating factors act as proxies for protected characteristics. Here is exact...
-
22 Mar 2026Build a CatBoost frequency-severity pricing model on freMTPL2 using Polars. Poisson frequency, Gamma severity, combined burning cost, SHAP factor extraction,...
-
21 Mar 2026Insurance walk-forward cross-validation prevents the look-ahead bias that makes standard k-fold results useless for prospective evaluation. Complete Python e...
-
21 Mar 2026sklearn's TweedieRegressor tutorial gets you to a fitted model in six lines. It also produces predictions that are wrong for any policy with non-annual expos...
-
21 Mar 2026NannyML is the best general-purpose ML monitoring library for teams without ground truth labels. For insurance pricing, it doesn't do exposure-weighted PSI, ...
-
21 Mar 2026Insurance model monitoring in Python that understands exposure weighting, development lags, and Gini drift. Why Evidently and NannyML miss what matters for p...
-
20 Mar 2026Standard GPD fitting is biased when claims are capped by policy limits. Most actuaries know this and do it anyway. insurance-severity v0.2.0 fixes it.
-
20 Mar 2026MAPIE is the standard Python library for conformal prediction, but it wasn't designed for insurance. Here is what goes wrong with exposure-weighted portfolio...
-
20 Mar 2026The FCA expects pricing teams to demonstrate their models don't proxy-discriminate under Consumer Duty. Most teams do this in Excel. Here is how to do it pro...
-
20 Mar 2026Most UK insurers don't hold ethnicity, religion, or disability status. Consumer Duty still requires evidence of fair outcomes. PrivatizedFairnessAudit solves...
-
20 Mar 2026Fairlearn is excellent for classification fairness. It was not built for insurance pricing, the Equality Act 2010, or the FCA's specific concern: proxy discr...
-
19 Mar 2026Single-quantile conformal methods apply one interval width across all forecast horizons. At 12 months ahead the interval is too narrow;
-
19 Mar 2026Active Consumer Duty investigations in home and travel insurance. What a defensible pricing model actually requires under PRIN 2A, and what the FCA thematic ...
-
19 Mar 2026Resampling your real portfolio to generate stress test scenarios destroys the joint tail structure that capital models depend on.
-
19 Mar 2026EquiPy is a technically excellent fairness correction tool built on optimal transport theory, from Arthur Charpentier's group at UQAM. insurance-fairness is ...
-
19 Mar 2026EconML is the standard Python library for causal ML. It was not built for insurance pricing, Poisson/Gamma exposure models, or the dual-selection bias proble...
-
18 Mar 2026Every UK pricing actuary smooths experience tables. Most do it with a 5-point moving average or a polynomial fitted by eye.
-
18 Mar 2026Pricing teams fit log-linear trends through experience with structural breaks. The straight line is wrong on both sides. How to detect and correct it.
-
18 Mar 2026DoWhy is the most rigorous general-purpose causal inference library in Python — DAG specification, formal identification, refutation tests. It was not built ...
-
18 Mar 2026Alibi Detect is a solid general-purpose drift detection library. It doesn't do exposure-weighted PSI, segmented A/E ratios, Gini discrimination drift, or Llo...
-
17 Mar 2026insurance-causal v0.3.1 fixes over-partialling in DML for small insurance books. Adaptive CatBoost regularisation makes causal estimates reliable at n≥1k.
-
17 Mar 2026Champion/challenger testing is the right way to evaluate pricing model changes. Most teams do it badly or not at all - ad-hoc scripts, no audit trail, no...
-
17 Mar 2026Static credibility weights all years equally. The dynamic Poisson-gamma state-space model weights recent experience more - and quantifies how much more.
-
16 Mar 2026Density ratio correction for portfolio composition shift - CatBoost classifier, importance-weighted evaluation, insurance-covariate-shift library
-
16 Mar 2026GLMComparison and MonotonicityEditor in insurance-gam close the governance gap between EBM shape functions and the GLM factor table your pricing committee...
-
16 Mar 2026Bootstrap and expert-judgment reserve ranges look like probability statements but carry no frequentist coverage guarantee.
-
15 Mar 2026Two things independent credibility cannot give you: a quantified uncertainty per sector, and stable factors year-on-year.
-
15 Mar 2026How to run covariate shift detection as a recurring monthly check: monitoring cadence, ESS ratio trends, and the thresholds that trigger a retraining...
-
15 Mar 2026Standard split conformal prediction requires exchangeability - a condition insurance claims time series systematically violate.
-
15 Mar 2026OLS elasticity in formula-rated books is contaminated by your own risk model. insurance-causal fixes this with CausalForestDML and CatBoost nuisance.
-
14 Mar 2026Automated GLM factor banding for UK insurance pricing: R2VF fused lasso, neural embeddings for high-cardinality categoricals, SKATER spatial clustering.
-
14 Mar 2026December is the season for year-end rate reviews where someone adds a flat 8% large loss loading to every segment regardless of tail weight.
-
14 Mar 202635 open-source Python libraries for UK insurance pricing: GBM-to-GLM distillation, causal inference, FCA fairness auditing, rate optimisation, PRA SS1/23.
-
14 Mar 2026insurance-governance merges insurance-validation and insurance-mrm. PRA SS1/23 statistical validation and MRM governance in one install - no version conflicts.
-
14 Mar 2026Three interpretable architectures for UK insurance pricing: EBM, ANAM, and PIN via insurance-gam. Refuse the GLM-vs-GBM accuracy trade-off with factor tables.
-
14 Mar 2026CausalForestDML separates causal price effect from risk-lapse correlation in UK motor renewal. insurance-elasticity - per-customer CATE and ENBP optimiser.
-
13 Mar 2026DiD and Callaway-Sant'Anna for rate change attribution. insurance-causal-policy quantifies what your rate change actually achieved, with FCA Consumer Duty-al...
-
13 Mar 2026Model risk governance for UK insurers: model inventory, risk tier scoring, exec committee reports. Built on SS1/23 principles (formally a banking standard, w...
-
13 Mar 2026Actuarial Neural Additive Models for UK pricing: exact interpretability, Tweedie loss, guaranteed monotonicity. insurance-gam - beyond SHAP and EBM limitations.
-
13 Mar 2026Champion/challenger with ICOBS 6B.2.51R compliance for UK insurers. SHA-256 routing, SQLite logging, bootstrap LR tests, SMF-signable report - insurance-deploy.
-
13 Mar 2026Two-stage CatBoost plus REML random effects for UK insurance broker adjustments. insurance-multilevel - Buhlmann-Straub credibility weighting, not guesswork.
-
13 Mar 2026Python insurance trend library: log-linear OLS/WLS, bootstrap CIs, ONS deflation, superimposed inflation, structural break detection - insurance-trend.
-
13 Mar 2026Continuous-time HMM for telematics risk scoring in UK motor pricing. Latent driving regimes from GPS data - actuarially interpretable features for Poisson GLM.
-
13 Mar 2026TabPFN and TabICLv2 for thin-segment UK insurance pricing. In-context learning at inference, no gradient descent. insurance-thin-data wraps both for actuaries.
-
13 Mar 2026Pairwise Interaction Networks produce exact tabulatable 2D rating factor surfaces, not SHAP approximations. Beats GBMs on French MTPL benchmark. Python.
-
13 Mar 2026Covariate-conditioned IBNR completion by risk segment using ML-EM algorithm. insurance-nowcast corrects aggregate LDF bias from your actual recent risk mix.
-
13 Mar 2026Joint conformal prediction sets for frequency and severity in UK insurance. Fan and Sesia coordinate-wise standardization - simultaneous coverage across both.
-
13 Mar 2026Four-tier experience rating in Python: Buhlmann-Straub, Poisson-Gamma state-space, GBM surrogate, attention credibility. Policy-level multiplicative factors.
-
13 Mar 2026EVT for UK motor large loss pricing: censored GPD for open TPBI claims, profile likelihood CIs, excess layer pure premiums. insurance-evt Python library.
-
13 Mar 2026Distributionally robust rate optimisation: worst-case demand within a Wasserstein ball. Price-of-robustness curve for UK pricing committee papers - Python.
-
13 Mar 2026Shared-trunk neural model for frequency-severity dependence in UK motor pricing. Explicit dependence testing where two-part GLMs assume independence - Python.
-
13 Mar 2026Conformal risk control for UK insurance: coverage calibrated to financial shortfall, not miscoverage rate. insurance-conformal - beyond standard intervals.
-
13 Mar 2026Spliced severity for UK motor BI: lognormal body, GPD tail above a policyholder-specific threshold. insurance-severity - mode-matching, 2,818 lines, 106 tests.
-
13 Mar 2026Bayesian Online Changepoint Detection for UK insurance loss ratios. Poisson-Gamma conjugates, regulatory event priors, Consumer Duty evidence pack - Python.
-
13 Mar 2026Logistic regression treats all non-lapsers the same. Mixture cure models split them into two groups: structural non-lapsers who will never leave, and...
-
12 Mar 2026Complete UK insurance pricing pipeline in Python: CatBoost GLM distillation, causal inference, FCA fairness auditing, rate optimisation, PRA SS1/23 governance.
-
12 Mar 2026Doubly robust TMLE for insurance pricing with Poisson outcomes and exposure offsets. insurance-tmle - first Python library with the implementation AIPW lacks.
-
12 Mar 2026Bandit algorithms for FCA GIPP-compliant price experimentation in UK general insurance. ENBP constraints and compliance reporting built in - insurance-online.
-
12 Mar 2026Neural Spline Flows for bimodal UK motor BI severity - no family assumption. insurance-nflow: TVaR, ILF curves, reinsurance layer costs, fat-tail transform.
-
12 Mar 2026Joint longitudinal-survival model for telematics: driving trajectory not current score. insurance-jlm - Wulfsohn-Tsiatis SREM with mid-term repricing in Python.
-
12 Mar 2026GARCH for UK insurance claims inflation: time-varying variance in trend analysis. insurance-garch - Engle (1982) applied to actuarial trend and pricing models.
-
12 Mar 2026EQRN covariate-dependent GPD tail modelling for XL pricing. Per-risk shape and scale via neural networks - insurance-quantile Python, Pasche and Engelke (2024).
-
12 Mar 2026Vine copulas for multi-peril UK home pricing. Flood-subsidence correlation costs ~9% in mispriced revenue. insurance-copula: BIC selection, PML simulation.
-
12 Mar 2026Fine-Gray subdistribution hazard for UK insurance competing risks. Separates lapse, MTC, and NTU correctly - insurance-survival Python, not naive censoring.
-
12 Mar 2026Causal Forests with Fixed Effects for UK insurance panel data. Rate change evaluation by segment - beyond before-and-after loss ratios. causalfe Python.
-
12 Mar 2026Bayesian Causal Forests for heterogeneous lapse effects in UK insurance pricing. Segment-level elasticity with posteriors - insurance-bcf wrapping stochtree.
-
12 Mar 2026Automatic Debiased ML via Riesz Representers for continuous price elasticity. insurance-causal - no GPS density blow-up at tails. UK personal lines Python.
-
12 Mar 2026Transfer learning for thin-segment UK insurance pricing: Tian-Feng GLM algorithm, CatBoost source-as-offset, CANN fine-tuning, negative transfer diagnostics.
-
11 Mar 2026MinTrace reconciliation for insurance pricing hierarchies: optimal joint adjustment across peril models and portfolio GLM. Exposure-weighted, Python.
-
11 Mar 2026Survival models for UK personal lines retention: cure models, survival-adjusted CLV, actuarial lapse tables, MLflow deployment. What lifelines does not do.
-
11 Mar 2026Regression Discontinuity Design tests if UK motor risk drops at age 25. Exposure-weighted Poisson outcomes, geographic boundaries, Consumer Duty output.
-
11 Mar 2026Double GLM gives every UK insurance policy its own dispersion parameter. insurance-dispersion - policy-level Solvency II variance and risk-adequate loading.
-
11 Mar 2026Mixture cure models for UK motor: separates non-claimers from susceptibles. Per-policyholder cure fraction scoring - insurance-survival Python library.
-
11 Mar 2026Detect threshold gaming in UK insurance using bunching estimators (Saez 2010, Kleven 2016). Mileage declarations, age spikes, sum-insured clustering - Python.
-
11 Mar 2026Credibility Transformer attention is mathematically identical to Buhlmann-Straub credibility. PyTorch, 1,746 parameters, beats CAFTT (27K). 85 tests.
-
10 Mar 2026GLMTransfer borrows statistical strength from a related source book to price thin target segments. Motor-to-fleet, home-to-landlord, and fleet roll-outs.
-
10 Mar 2026A 12% rate increase on young motor drivers. An 8% lapse spike three months later. Here is how to tell whether the rate change caused it — using synthetic dif...
-
10 Mar 2026Discrimination-free UK insurance pricing via Wasserstein barycenter and causal path decomposition. Satisfies FCA Consumer Duty proxy discrimination rules.
-
10 Mar 2026GAMLSS in Python: seven families, RS algorithm, variance as function of covariates. insurance-distributional-glm - the actuarial implementation Python lacked.
-
10 Mar 2026Distributional Refinement Networks wrap any GLM to produce a full predictive distribution. insurance-severity - neural severity modelling for UK motor pricing.
-
09 Mar 2026Whittaker-Henderson smoothing for noisy experience rating tables in Python. REML lambda selection, Bayesian confidence intervals, 2D surface smoothing.
-
09 Mar 2026From CatBoost frequency model to BYM2 spatial territory factors for Emblem or Radar. Data engineering, MCMC convergence checks, Polars joins - Python.
-
09 Mar 2026Handle 800+ vehicle makes and 9,000+ postcode sectors in a multiplicative GLM using neural embeddings and spatial clustering. Auditable Python pipeline.
-
09 Mar 2026Actuarially faithful synthetic data via vine copulas and AIC-selected marginals. insurance-synthetic fixes Poisson semantics and tail behaviour SDV gets wrong.
-
09 Mar 2026The full calibration framework for insurance pricing: balance property test, auto-calibration by price cohort, and Murphy decomposition...
-
09 Mar 2026insurance-gam wraps EBM for UK pricing teams: Poisson/Tweedie loss, exposure offsets, RelativitiesTable, MonotonicityEditor, GLM comparison diagnostics.
-
09 Mar 2026Where double machine learning beats naive regression for insurance pricing — and where it does not. Benchmarks on 100,000-policy synthetic UK motor data with...
-
08 Mar 2026Three diagnostics prove a GAMLSS sigma submodel is real: quantile residuals, worm plots, split-sample calibration. From insurance-distributional-glm.
-
08 Mar 2026GAS filters track claims frequency and severity trend between GLM refits. Step-by-step tutorial using insurance-gas on UK motor data.
-
07 Mar 2026CatBoost MultiQuantile plus actuarial output layer: TVaR, ILFs, large loss loadings, exceedance probabilities for UK insurance pricing. insurance-quantile.
-
07 Mar 2026Constrained portfolio rate optimiser for UK personal lines: SLSQP, analytical Jacobians, FCA ENBP enforcement, shadow prices. Python - insurance-optimise.
-
06 Mar 2026ICC diagnostics for multiple group factors in insurance pricing. When broker, scheme, fleet, and postcode sector effects are worth modelling with REML...
-
06 Mar 2026When a new aggregator partnership or competitor exit changes your new business mix, models trained on the old distribution misprice silently.
-
05 Mar 2026insurance-distributional models the full conditional loss distribution, not just the mean. First open-source Python implementation of the ASTIN 2024 Best Paper.
-
04 Mar 2026Per-risk large loss loadings for UK home insurance using quantile GBMs. Avoids the flat-loading trap by making the loading a function of the risk itself.
-
04 Mar 2026Step-by-step tutorial: plant two interactions in synthetic motor data, detect them with CANN + NID, validate with SHAP, confirm with A/E surfaces, and...
-
03 Mar 2026Detect and correct proxy discrimination in UK insurance using SHAP and insurance-fairness. Protected characteristic leakage detection under FCA Consumer Duty.
-
03 Mar 2026Three-layer drift detection: feature drift, segmented calibration, Gini test. Tells you whether to recalibrate or refit - beyond PSI and A/E ratios.
-
02 Mar 2026The foundational walkthrough for insurance-covariate-shift: density ratio estimation, ESS/KL diagnostics, importance weighting, shift-robust conformal...
-
02 Mar 2026Step-by-step: extract multiplicative CatBoost rating factors using shap-relativities. SHAP decomposition to GLM-format exp(beta) tables with CI and...
-
01 Mar 2026Double Machine Learning fixes biased price elasticity in insurance quote data. insurance-optimise: conversion, retention, elasticity, demand curves, FCA GIPP.
-
01 Mar 2026Python library distilling CatBoost GBMs into multiplicative GLM factor tables for Radar and Emblem. Open-source GBM-to-GLM distillation for UK pricing teams.
February 2026
12 articles
-
28 Feb 2026How to combine GLM and GBM predictions for production pricing: cross-validated blend weights, PRA interpretability, and when blending actually helps. Once th...
-
28 Feb 2026Bühlmann-Straub vs CatBoost vs two-stage multilevel for UK motor pricing: when each wins and how insurance-credibility and insurance-multilevel combine them.
-
28 Feb 2026Assumes familiarity with the Murphy decomposition framework. Focuses on the operational question: given a monitoring alert, how do you read GMCB vs LMCB...
-
27 Feb 2026Automated interaction search for UK motor GLMs using CANN residuals and NID. Bonferroni-corrected shortlist before manual testing - insurance-interactions.
-
27 Feb 2026Python library for NCD and bonus-malus in UK motor insurance. Optimal claiming thresholds peak at 20% NCD discount, not 65% - derived mathematically.
-
23 Feb 2026BYM2 spatial model in PyMC for UK territory ratemaking. Borrows strength across neighbouring postcode sectors - statistically correct vs k-means banding.
-
21 Feb 2026Databricks workflow for UK pricing actuaries: CatBoost plus MLflow tracking, SHAP relativities, and Radar export. End-to-end motor pricing in Python.
-
21 Feb 2026Rate changes that meet a target loss ratio, respect movement caps, and minimise cross-subsidy. Linear programming for UK personal lines pricing teams.
-
19 Feb 2026Distribution-free conformal prediction intervals for insurance GBMs. Per-risk coverage guarantees, not confidence intervals for the mean. Python library.
-
19 Feb 2026Buhlmann-Straub credibility in Python for UK personal lines. Blend thin-segment experience with portfolio rates - mathematically equivalent to mixed models.
-
17 Feb 2026How to extract SHAP relativities from insurance GBMs. Multiplicative factor tables in GLM exp(beta) format, with confidence intervals and exposure weighting....
-
17 Feb 2026Partial pooling for thin rating cells in UK motor pricing. bayesian-pricing stabilises sparse segments with hierarchical Bayesian models - no data discarded.
July 2025
2 articles
-
28 Jul 2025How to convert raw telematics trip data into GLM-ready features for UK motor pricing. Covers HMM state segmentation and score calibration to GLM relativities.
-
13 Jul 2025PRA SS1/23 requires quantitative pass/fail tests, not narrative. insurance-governance automates the full validation suite and generates auditable HTML reports.
May 2025
1 articles
-
14 May 2025Your frequency GLM and severity GLM are both correct. Multiplying them is not. How to test and correct for the dependence your pricing model ignores.
March 2025
1 articles
-
15 Mar 2025A practitioner tutorial on fitting spliced composite severity distributions for UK motor claims using insurance-severity.