page-banner

Tackling the hardest
problem in capital modelling

Insurance

How we helped bridge the gap between the business view of risk drivers and the allowance for dependencies in internal models.

The background

Our client’s internal model covers a variety of risks at a very granular level. How these risks are combined is effectively the most significant assumption in the model, and impacts the capital requirements and results used for business decisions, such as reinsurance strategy.

Like many insurers, our client had been grappling with how to get comfortable with the dependency assumptions, and particularly with how to demonstrate objectively that the assumptions were consistent with management’s view of the business and its unique risk profile.

Our solution

We developed ways of assessing the modelling of dependencies against the business view of risk. This involved:

  • Workshops with stakeholders to identify the key drivers of risk and capture their views on the impact of each driver on the business
  • Using these views to parameterise and run an independent model (LCP’s Dependency Scenario Generator) to combine the effects of each driver on the relevant claims models, and to assess the dependency of the underwriting and reserving risk
  • Comparing the independent results with the main internal model results and identifying areas of similarity and difference.

The results

Our analysis suggested that, although the capital number was broadly appropriate, the downside risk at the 1/10 probability was most likely being overstated by the internal model.

This confirmed what both management and the capital team had intuitively felt for some time. It enabled them to quantify the effect and to begin investigating options for refining the model to make it more appropriate for day-to-day use.

How we can help

We work with insurers to help them better understand and manage the risks they face and their capital requirements.

Our business-focused approach to validation can help you be confident that your insurance models are robust, and that you are meeting regulatory requirements.

Want more information?

If you would like to know more about our services and how we can help you with insurance

Contact us