Summary Claims, reserving and model risk are usually organised as separate disciplines. In practice they form one chain: claims experience informs reserving assumptions, reserving methodology is embedded in models, and model outputs shape pricing, capital and strategy. This article sets out how to govern the chain rather than the links.
Claims governance Claims handling quality shapes both customer outcomes and financial outcomes. Governance covers case reserve accuracy, settlement discipline, leakage indicators, and fraud controls. Weak claims data undermines everything downstream.
Reserving methodology Reserving methodology should be documented, validated, and subject to independent review. Best-estimate reserves are accompanied by uncertainty ranges, management overlays are governed, and reserving results are challenged by the risk function and, where material, by external actuaries.
Model risk Reserving models, pricing models, capital models and increasingly machine-learning models used in claims triage and fraud detection all require model risk governance: inventory, tiering, independent validation, ongoing monitoring, and change control. The insurance context introduces long-tailed liabilities where validation cannot rely on short-horizon back-testing alone.
The chain Weakness at any point cascades. Poor case reserves distort development triangles, which distort reserving assumptions, which distort model calibrations, which distort pricing and capital decisions. Governance should trace the chain end to end, not audit each link in isolation.
CRO and board implications Boards should see how claims quality, reserving stability and model performance interact — for example, a year in which reserving strengthens materially should prompt questions about pricing and appetite implications, not only about the reserving decision itself.
Practical implementation Claims MI tied into reserving governance; documented reserving methodology with uncertainty ranges and overlay register; model risk framework covering all material actuarial and analytical models; annual joint review across claims, actuarial and risk functions.
Limitations Insurance liabilities are inherently uncertain over long horizons. Governance discipline manages that uncertainty; it does not eliminate the possibility of adverse development.
Related reading See [Insurance Risk & Solvency](/expertise/insurance-risk), [Model Risk](/expertise/model-risk) and [Enterprise Risk](/expertise/enterprise-risk).
Frequently asked questions
What should risk leaders know about claims governance?
Claims handling quality shapes both customer outcomes and financial outcomes. Governance covers case reserve accuracy, settlement discipline, leakage indicators, and fraud controls. Weak claims data undermines everything downstream.
What should risk leaders know about reserving methodology?
Reserving methodology should be documented, validated, and subject to independent review. Best-estimate reserves are accompanied by uncertainty ranges, management overlays are governed, and reserving results are challenged by the risk function and, where material, by external actuaries.
What should risk leaders know about model risk?
Reserving models, pricing models, capital models and increasingly machine-learning models used in claims triage and fraud detection all require model risk governance: inventory, tiering, independent validation, ongoing monitoring, and change control. The insurance context introduces long-tailed liabilities where validation cannot rely on short-horizon back-testing alone.
What should risk leaders know about the chain?
Weakness at any point cascades. Poor case reserves distort development triangles, which distort reserving assumptions, which distort model calibrations, which distort pricing and capital decisions. Governance should trace the chain end to end, not audit each link in isolation.
What should risk leaders know about cRO and board implications?
Boards should see how claims quality, reserving stability and model performance interact — for example, a year in which reserving strengthens materially should prompt questions about pricing and appetite implications, not only about the reserving decision itself.