Insurance & Climate

Solvency and capital modelling in insurers

Design and validation considerations for internal and standard-formula capital models.

By Jonas Adam Mohamed Osman AbdelghafourPublished 15 November 2025

Summary

Capital models — internal or standard-formula — are the anchor of solvency governance. Design and validation decisions determine whether they inform decisions or merely satisfy filings.

Internal vs standard formula

Internal models offer better fit to specific risk profiles and greater governance burden. Standard formula is simpler and often over- or under-conservative at the tails. The choice is strategic.

Model use

Regulatory use tests require that the model informs actual decisions: pricing, reinsurance, capital allocation, strategy. Use that is documented but not observable is a finding waiting to happen.

Validation

Validation covers data, methodology, assumptions, calibration, results and governance. It should include stability analysis and challenge of key assumptions such as correlations.

Reinsurance modelling

Reinsurance is a material lever for capital. Modelling should reflect contract terms accurately, including reinstatements, aggregate limits and counterparty risk.

Governance

The board should see the model's material limitations and the compensating controls, not only headline coverage ratios.

Limitations

Capital models are point-in-time constructions calibrated to the observable past. Tail risks by nature test the extrapolation.

Related expertise

See Insurance Risk & Solvency and Model Risk and Validation.

Frequently asked questions

What should risk leaders know about internal vs standard formula?

Internal models offer better fit to specific risk profiles and greater governance burden. Standard formula is simpler and often over- or under-conservative at the tails. The choice is strategic.

What should risk leaders know about model use?

Regulatory use tests require that the model informs actual decisions: pricing, reinsurance, capital allocation, strategy. Use that is documented but not observable is a finding waiting to happen.

What should risk leaders know about validation?

Validation covers data, methodology, assumptions, calibration, results and governance. It should include stability analysis and challenge of key assumptions such as correlations.

What should risk leaders know about reinsurance modelling?

Reinsurance is a material lever for capital. Modelling should reflect contract terms accurately, including reinstatements, aggregate limits and counterparty risk.

What should risk leaders know about governance?

The board should see the model's material limitations and the compensating controls, not only headline coverage ratios.