Actuarial-Risk-modeling-icone

In the Solvency II regulation framework, a model can be approved for the computation of capital requirements if it respects the condition of being validated by the undertaking.

The validation step of such a model gives some confidence to the company that uses it.

Various model components (tools, programs, processes and quality standards) can be subject to a validation procedure: methods, data quality, assumptions, expert judgment, documentation, IT systems, model governance, use test,

Consequently, the validation can be defined as a set of tools and processes (both quantitative and qualitative) used by the company to gain confidence about the results, concepts, processes and work carried out as part of the model. The model validation scope should not be restricted to the core calculation engine to obtain the capital requirements, but should also include all quantitative and qualitative processes involved in the model.

Solvency II Directive

Compliance with Directive 2009/138/EC imposes regular and independent cycle of model validation. A fully elaborated model validation policy is also required  including at least a detailed description of its mission statement, its scope, the tasks and roles with respect to internal and external validation teams, its resources, its reporting lines, its tools and its methodology.

Legal documentation can help building the validation framework:

  • Level of compliance about Use test (Article 120)
  • Statistical quality standards (Article 121)
  • Calibration standards (Article 122)
  • Profit and loss attribution (Article 123)
  • Validation (Article 124)
  • Documentation standards (Article 125)

A project of model validation can be organized in five main streams:

1. (Economic) valuation model

2. Cash-flows modelling

3. Economic Scenarios Generator

4. Capital requirements calculation: SCR, MCR, …

5. Own Risk and Solvency Assessment (ORSA): risk appetite, …