The Credibility Report
Actuarial Intelligence for Insurance Professionals
What’s in this edition
Primary-source market updates (no aggregator links) plus the latest actuarial-relevant arXiv papers (score ≥ 15, last 14 days).
📰 Headlines (primary sources)
Catastrophe Bonds: An Uncorrelated Asset Class Amid Global Macroeconomic Uncertainty - Neuberger Berman
Read source → • Neuberger BermanFrom risk transfer to risk prevention: How AI supports long-term financial resilience in insurance - Microsoft
Read source → • MicrosoftHow the insurance sector is stepping up on climate adaptation and resilience - Environmental Defense Fund
To build a climate-resilient future, we need to understand what adaptation and resilience measures insurers are driving, and how to scale what works.
Read source → • Environmental Defense FundTexas Universities Show Sharp Differences in Health Insurance Premiums - Baker Institute
Read source → • Baker InstituteHow provider-led health plans can succeed in commercial insurance - McKinsey & Company
Read source → • McKinsey & CompanyReConnect 2026 Showcases Cayman’s Expanding Role in Global Reinsurance - AM Best
Read source → • AM Best🔬 Research Spotlight (arXiv)
Your SaaS Is an Insurance Product: A Modeling Framework
arXiv • Score: 43 • 2026-05-15
Capped-usage SaaS products -- LLM subscriptions such as Claude Code and ChatGPT, cloud platforms such as Vercel and Cloudflare Workers, corporate benefit platforms, identity-verification services with liability transfer -- share a structural signature with insurance products: a fixed premium decoupled from realized consumption, stochastic per-user demand with heavy-tailed severity, a non-fungible cap that resets on a fixed schedule, and a portfolio-level exposure that requires reserve adequacy under tail risk. We argue that this is not an analogy. It is the same operational problem actuarial science has been tooled for decades to address, restated with new dependent variables (tokens, bandwidth bytes, function-invocations, gym check-ins) in place of medical claims. This paper proposes a modeling framework for capped-usage SaaS pricing built from frequency-severity decomposition, premium calculation principles, and Monte Carlo reserve adequacy. We map the framework to publicly observable subscription tiers in two domains (LLM services and cloud platforms), ground it in canonical health-insurance economics (Arrow 1963; Pauly 1968; Manning et al. 1987; Brot-Goldberg et al. 2017), and demonstrate divergence from traditional unit economics through a worked example. The contribution is operational rather than theoretical: not a new theorem, but vocabulary and tools currently absent from cs.LG/stat.ML practice.
Open paper →Fairness Testing for Algorithmic Pricing
arXiv • Score: 42 • 2026-05-12
Algorithmic systems now set prices across auto insurance, credit, and lending markets, and regulators increasingly require firms to demonstrate that these systems do not discriminate against protected groups. The standard audit regresses pricing output on a protected attribute and legitimate rating factors, then tests the resulting coefficient using ordinary least squares standard errors. We show that this approach is structurally invalid. Pricing algorithms are usually deterministic, so residuals reflect approximation error rather than sampling variability, rendering classical standard errors invalid in both direction and magnitude. We derive correct asymptotic variance estimators for OLS and GLM audit regressions and the correct cross-covariance formula for proxy discrimination testing. Applied to quoted premiums from 34 Illinois auto insurers, every insurer fails the conditional demographic parity test, with minority zip codes paying $34-$158 more per year than comparable-risk white zip codes. The standard proxy discrimination formula flags zero insurers. However, our corrected formula identifies all 34 as statistically significant, of which 16 exceed the substantive threshold. Our framework provides statistically valid audit tools for any deterministic algorithmic system subject to regression-based fairness testing.
Open paper →Mortality Heterogeneity and Actuarial Fairness in China's Notional Defined Contribution Pension System
arXiv • Score: 25 • 2026-05-18
We study actuarial fairness in China's notional defined contribution (NDC) pension system when mortality differs across income groups. Under current rules, individual account balances are converted into monthly benefits using an official annuity divisor that depends only on retirement age. We develop a mortality-differentiated Lee-Carter framework with group-specific baseline mortality schedules and a common period effect, estimated by combining national mortality data for 1994-2020 with CHARLS subgroup data for 2011-2020. To model cross-group mortality parsimoniously under limited data, we parameterize the baseline schedules using Hermite splines. Applying the model to China's NDC system, we find substantial actuarial unfairness in the current age-only divisor. The subsidy rises monotonically with income, implying both an aggregate actuarial shortfall and a reverse transfer from poorer to richer retirees. We then compare four implementable income-dependent annuitization rules, ranging from a simple bracket design to marginal-rule alternatives, and show that all substantially reduce the reverse transfer.
Open paper →A Model-Agnostic Bootstrap for Macro-Level Claims Reserving Under the Conditioning Principle
arXiv • Score: 22 • 2026-05-15
The correct inferential object in claims reserving is the conditional predictive distribution $p(R \mid \mathcal{D}, \hatθ)$, where $\mathcal{D}$ is the observed triangle held fixed. We refer to this as the conditioning principle. All existing bootstraps violate it by resampling functions of $\mathcal{D}$ inside the predictive loop, producing an $O(1)$ coverage error that does not vanish as the triangle grows. The Dirichlet-Gamma hierarchy admits a bootstrap that satisfies the principle exactly: $S^{IBNP}_i = X^{obs}_i (1-W_i)/W_i$ with $W_i \sim \mathrm{Beta}(c\hat{F}_{I-i}, c(1-\hat{F}_{I-i}))$ sampled directly from its predictive distribution. Only the allocation proportion $W_i$ is simulated; the observed triangle is held fixed. It thus inherits calibration from any development-proportion method (Chain-Ladder, Bornhuetter-Ferguson, Cape Cod, or other), making it model-agnostic. The coverage deficit is $O(I^{-1/2})$, independent of the number of development periods. Under compound Poisson data-generating processes the bootstrap is conservative for every $F_{I-i} \in (0,1)$: the predictive standard deviation analytically exceeds the true value by the factor $1/\sqrt{F_{I-i}}$. The ODP bootstrap violates the principle through two mechanisms in opposite directions: re-estimation inflates bootstrap variance under the ODP DGP, while missing accident-year frailty deflates it under frailty DGPs. The resulting coverage discrepancy is $Ω(1)$ regardless of $I$, providing a structural explanation for the cross-portfolio miscalibration heterogeneity documented by Meyers (2015). Chain-Ladder, Bornhuetter-Ferguson and Cape Cod emerge as credibility estimators under diffuse, informative and pooling priors respectively, with identical structure for counts and amounts. The concentration $c$ serves as a diagnostic: $\hat{c} < 30$ signals non-stationary development.
Open paper →The Negative Binomial Chain-Ladder: A Full Likelihood Model for Claim Count Reserving
arXiv • Score: 20 • 2026-05-15
The Chain-Ladder (CL) method remains the dominant macro-level technique for claims reserving in non-life insurance, yet its classical formulation lacks a coherent probabilistic foundation. Existing stochastic extensions-including the Mack model and the Over-Dispersed Poisson (ODP) framework-provide measures of uncertainty but rely on second-moment assumptions or quasi-likelihood variance structures without clear generative interpretations. This paper develops a Negative Binomial Chain-Ladder (NB-CL) model that embeds the CL method within a full likelihood-based framework. The key contribution is a micro-level derivation showing that the negative binomial distribution arises naturally from a Poisson-Gamma construction: claims arrive according to a Poisson process with Gamma-distributed accident-year heterogeneity, and aggregation yields negative binomial incremental counts. This derivation gives the dispersion parameter $κ$ a structural interpretation as accident-year heterogeneity, rather than an ad-hoc overdispersion adjustment. The NB-CL model generalises the Poisson Chain-Ladder model in the limit $κ\to \infty$, shares the point estimates of the ODP model while differing in its variance function (quadratic vs. linear), and unifies the Chain-Ladder family within a single probabilistic hierarchy. A parametric bootstrap procedure is developed to incorporate both process and parameter uncertainty. Simulation studies confirm near-nominal coverage under correct specification once the dispersion parameter is bias-corrected, and a controlled degradation under model misspecification. Empirical illustrations on claim count data (Australian motor bodily injury) and paid amounts (Taylor-Ashe) document both the structural reading of $κ$ and the working-approximation status of the model in the amounts case.
Open paper →Asymptotic Behaviour of Unexpected Losses and Risk Ratios for Co-Monotonic Alternatives
arXiv • Score: 17 • 2026-05-18
The aggregation of individual risks in large credit and insurance portfolios is guided by diversification and the law of large numbers, which formalizes the convergence of sample averages to their means. At the same time, regulatory capital requirements and insurance premia are designed to provide a capital buffer or risk margin above the mean. The resulting excess, given by the difference between the nonlinear valuation of the aggregate loss and the corresponding mean, reflects the idea of protection against unexpected losses in the sense of banking and insurance regulation. This paper studies the asymptotic behaviour of this excess for large weighted portfolios. The main result shows that, for monotone cash-additive risk measures on Banach-lattice-valued Orlicz spaces, convergence along weighted averages satisfying a weak law of large numbers together with a uniform integrability condition is equivalent to scalar continuity at the origin. If the risk measure is positively homogeneous, this continuity condition is automatically satisfied, and we prove that the unexpected losses of large weighted portfolios are of order $o(n\overlineλ_n)$, where $\overlineλ_n$ denotes the average weight assigned to the first $n$ random variables. We establish analogous asymptotic results for Choquet insurance premia. Finally, we derive risk-ratio limits that quantify the potential underestimation arising when diversified portfolios are compared with co-monotonic alternatives.
Open paper →✅ Practical Takeaways
- For P&C pricing and capital work, refresh wildfire and severe-convective-storm accumulation scenarios rather than relying on last year’s peril mix.
- For property underwriting, test whether post-wildfire resilient rebuilding standards justify explicit mitigation credits or revised rebuild-cost assumptions.
- For MTPL frequency models, benchmark zone-level coordinates and environmental features against the existing tariff variables before adding more complex image embeddings.
- For health insurance valuation, run stochastic inflation and interest-rate sensitivity alongside deterministic best-estimate calculations.
Until next time—stay credible.
— The Credibility Report
Edition 021 | Prepared May 24, 2026 (UTC)